From e4968064b0ef642808cfb2eaa158f8ee0bd8356b Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Thu, 20 Nov 2025 00:12:34 +0100 Subject: [PATCH 01/18] New translations panelcleaner_source.ts (Spanish) --- translations/PanelCleaner_es_ES.ts | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/translations/PanelCleaner_es_ES.ts b/translations/PanelCleaner_es_ES.ts index c60701c..8fd07e9 100644 --- a/translations/PanelCleaner_es_ES.ts +++ b/translations/PanelCleaner_es_ES.ts @@ -2788,7 +2788,7 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con Pause Countdown - Pause Countdown + Pausar cuenta regresiva @@ -4574,7 +4574,7 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Romanian; Moldavian Language code option - Romanian; Moldavian + Rumano; Moldavo @@ -4681,7 +4681,7 @@ Puede descargar el modelo manualmente desde <a href="https://github.com/ Regex is valid - Regex is valid + Regex es válido From bada6fe58f957e21b24bec55e6bcff2523054b2b Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Thu, 20 Nov 2025 02:03:29 +0100 Subject: [PATCH 02/18] New translations panelcleaner_source.ts (Spanish) --- translations/PanelCleaner_es_ES.ts | 386 ++++++++++++++--------------- 1 file changed, 193 insertions(+), 193 deletions(-) diff --git a/translations/PanelCleaner_es_ES.ts b/translations/PanelCleaner_es_ES.ts index 8fd07e9..fa39b8e 100644 --- a/translations/PanelCleaner_es_ES.ts +++ b/translations/PanelCleaner_es_ES.ts @@ -196,7 +196,7 @@ Burbujas eliminadas: Total boxes - Total cajas + Total de cajas @@ -549,27 +549,27 @@ Páginas con fallos / total: %id - %id + %id %i - %i + %i %od - %od + %od %o - %o + %o %p - %p + %p @@ -2277,7 +2277,7 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con Review - Review + Revisar @@ -2590,167 +2590,167 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con <html><head/><body><p>If you set a wait time greater than 0, a window will open upon process completion. This window will show a countdown and allow you to preview as well as edit the command.</p></body></html> - <html><head/><body><p>If you set a wait time greater than 0, a window will open upon process completion. This window will show a countdown and allow you to preview as well as edit the command.</p></body></html> + <html><head/><body><p>Si establece un tiempo de espera superior a 0, se abrirá una ventana al finalizar el proceso. Esta ventana mostrará una cuenta atrás y le permitirá previsualizar y editar el comando.</p></body></html> Power Off Command: - Power Off Command: + Comando de apagado: shutdown - shutdown + apagar Custom Commands: - Custom Commands: + Comandos personalizados Move command up - Move command up + Mover comando hacia arriba Move command down - Move command down + Mover comando hacia abajo New - New + Nuevo Delete - Delete + Eliminar Action Name - Action Name + Nombre de la acción Command or Script File Path - Command or Script File Path + Ruta del archivo de comandos o scripts Cancel the custom action if processing failed - Cancel the custom action if processing failed + Cancelar la acción personalizada si falla el procesamiento <html><head/><body><p>The Power Off action will run regardless of this setting.</p></body></html> - <html><head/><body><p>The Power Off action will run regardless of this setting.</p></body></html> + <html><head/><body><p>La acción de apagado se ejecutará independientemente de esta configuración.</p></body></html> Placeholder Tokens: - Placeholder Tokens: + Tokens de marcador de posición Input Files - Input Files + Archivos de entrada %i Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. - %i + %i <html><head/><body><p>The individual file paths of each file processed, as a space separated list, with quotation marks where necessary.</p></body></html> - <html><head/><body><p>The individual file paths of each file processed, as a space separated list, with quotation marks where necessary.</p></body></html> + <html><head/><body><p>Las rutas individuales de cada archivo procesado, como una lista separada por espacios, con comillas donde sea necesario.</p></body></html> Input Directory - Input Directory + Directorio de entradas %id Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. - %id + %id <html><head/><body><p>Group files by parent directory and list these, as a space separated list, with quotation marks where necessary.</p></body></html> - <html><head/><body><p>Group files by parent directory and list these, as a space separated list, with quotation marks where necessary.</p></body></html> + <html><head/><body><p>Agrupe los archivos por directorio principal y enumérelos, como una lista separada por espacios, con comillas donde sea necesario.</p></body></html> Output Files - Output Files + Archivos de Salida %o Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. - %o + %o <html><head/><body><p>The individual file paths of each file created as an output, as a space separated list, with quotation marks where necessary.</p></body></html> - <html><head/><body><p>The individual file paths of each file created as an output, as a space separated list, with quotation marks where necessary.</p></body></html> + <html><head/><body><p>Las rutas individuales de cada archivo creado como salida, como una lista separada por espacios, con comillas donde sea necesario.</p></body></html> Output Directory - Output Directory + Directorio de salida %od Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. - %od + %od <html><head/><body><p>Group output files by parent directory and list these, as a space separated list, with quotation marks where necessary.</p></body></html> - <html><head/><body><p>Group output files by parent directory and list these, as a space separated list, with quotation marks where necessary.</p></body></html> + <html><head/><body><p>Agrupe los archivos de salida por directorio principal y enumérelos, como una lista separada por espacios, con comillas donde sea necesario.</p></body></html> Profile Used - Profile Used + Perfil utilizado %p Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. - %p + %p <html><head/><body><p>The name of the profile used, with quotation marks if necessary.</p></body></html> - <html><head/><body><p>The name of the profile used, with quotation marks if necessary.</p></body></html> + <html><head/><body><p>El nombre del perfil utilizado, entre comillas si es necesario.</p></body></html> Action names cannot be empty. - Action names cannot be empty. + Los nombres de las acciones no pueden estar vacíos. Action names must be unique. - Action names must be unique. + Los nombres de las acciones deben ser únicos. The name "{shutdown_command}" is reserved. - The name "{shutdown_command}" is reserved. + El nombre "{shutdown_command}" está reservado. The Flatpak sandbox will likely prevent actions from working. - The Flatpak sandbox will likely prevent actions from working. + Es probable que el sandbox de Flatpak impida que las acciones funcionen. @@ -2758,32 +2758,32 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con Post-Action Runner - Post-Action Runner + Ejecutador de Post-Action An error occurred during the process. - An error occurred during the process. + Ocurrió un error durante el proceso. Full Command: - Full Command: + Comando completo: Run Again - Run Again + Ejecutar de nuevo Skip Countdown - Skip Countdown + Saltar cuenta regresiva Resume Countdown - Resume Countdown + Reanudar cuenta regresiva @@ -2798,32 +2798,32 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con Cancel Action - Cancel Action + Cancelar acción Cancel Shutdown - Cancel Shutdown + Cancelar Apagado Shutting down the system - Shutting down the system + Cerrando el sistema. Running the action "{action_name}" - Running the action "{action_name}" + Ejecutar la acción "{action_name}" Remaining time: {time} seconds - Remaining time: {time} seconds + Tiempo restante: {time} segundos Remaining time: 1 second - Remaining time: 1 second + Tiempo restante: 1 segundo @@ -3074,13 +3074,13 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output - <This is just a placeholder, translating this string is not necessary> + <Esto es solo un marcador de posición, la traducción de esta cadena no es necesario> Write Output Output name in the image details view - Write Output + Salida de escritura @@ -3167,73 +3167,73 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General - Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). + Divida tiras largas en páginas individuales. Si se activa, en lugar de apretar la tira entera para caber en la altura preferida, el clip se dividirá en páginas individuales, cada ajuste de la altura preferida más/menos doble el margen de tolerancia dividida (si el segmento estaba en medio de la raya). Split Long Strips Profile entry in General - Split Long Strips + Dividir Tiras Largas Preferred height to split long strips at. Profile comment in General - Preferred height to split long strips at. + Altura preferida para dividir tiras largas. Preferred Split Height Profile entry in General - Preferred Split Height + Altura de división preferida Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General - Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. + Margen de tolerancia para dividir tiras largas, es la diferencia máxima entre la altura de división preferida y la altura real de la raya. Un algoritmo determina el mejor punto de división dentro de este margen, en un esfuerzo por evitar dividirse en medio de un panel. Split Tolerance Margin Profile entry in General - Split Tolerance Margin + Margen de tolerancia de división Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General - Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. + Relación de aspecto a utilizar para dividir tiras largas. Es la relación entre la anchura y la altura de la imagen. Si la relación de aspecto de la imagen'es menor que este valor, se considerará una tira larga.<br> Ejemplo: 0,3 significa que la anchura de la imagen es 0,3 veces la altura. Long Strip Aspect Ratio Profile entry in General - Long Strip Aspect Ratio + Relación de aspecto de Tira larga Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General - Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. + Combinar tiras largas en una sola imagen. Si se activa, las páginas individuales creadas a partir de un clip largo se fusionarán de nuevo en una sola imagen al exportar. Merge After Split Profile entry in General - Merge After Split + Fusionar después de dividir Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General - Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + Número máximo de hilos a usar para exportar imágenes. Puede dejarlo sin especificar para utilizar todos los hilos disponibles. Reduzca este valor si tiene problemas de memoria, que aparecerán como bloqueos aleatorios. Max Threads Export Profile entry in General - Max Threads Export + Exportación máxima de hilos @@ -3343,13 +3343,13 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor - The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. + El idioma a utilizar para las tareas de OCR. El detector de texto puede detectar japonés e inglés, para cualquier otra cosa, seleccione un idioma explícitamente aquí. La detección por cuadro retiene lo que el detector de texto detectó, mientras que por página asignará el idioma más prominente a toda la página.<br> Nota: Sólo el japonés es compatible de fábrica, todo lo demás requiere que Tesseract esté habilitado y que se instalen sus paquetes de idiomas asociados. OCR Language Profile entry in Preprocessor - OCR Language + Idioma OCR @@ -3392,7 +3392,7 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor - Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. + Patrón Regex para comparar con los resultados del OCR. Todo lo que coincida con este patrón se descarta. Tenga en cuenta que el modelo OCR devuelve caracteres de ancho completo, por lo que este patrón debería coincidir con ellos. @@ -3405,7 +3405,7 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor - The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. + El modelo OCR sólo puede manejar texto en japonés, por lo que cuando se activa la opción estricta, descartará las casillas que no'confíe que son japonesas. A veces, los números u otros símbolos disminuyen su confianza, lo que hace que el idioma detectado sea desconocido. Si se desactiva la opción estricta, no se descartarán. Todo lo que se reconozca con seguridad como un idioma diferente se descartará independientemente de esta configuración. @@ -3478,13 +3478,13 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker - Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + Número máximo de hilos a usar para la generación de máscaras. Puede dejarlo sin especificar para utilizar todos los hilos disponibles. Bajar este valor si se encuentran problemas de memoria, que aparecerán como bloqueos aleatorios. Max Threads Profile entry in Masker - Max Threads + Hilos máximos @@ -3526,37 +3526,37 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter - The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. + La desviación estándar mínima de colores alrededor del borde de una máscara determinada para realizar pintura en la región alrededor de la máscara. Si este valor coincide con la desviación máxima para las máscaras, sólo se pintarán burbujas fallidas, haciendo que los siguientes dos ajustes sean irrelevantes. Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General - Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. + Combine las salidas en un solo archivo de proyecto como capas. Formatos soportados actualmente: Photoshop PSD.<br> - ninguno: Cada imagen y máscara se guardan como archivos básicos.<br> - psd-per-imagen: Las imágenes y máscaras se guardan juntas en un archivo PSD por imagen de entrada.<br> - psd-bulk: Todas las imágenes y máscaras se guardan juntas en un único archivo PSD, agrupado por imagen de entrada. Layered Export Profile entry in General - Layered Export + Exportación por capas Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker - Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. + Si se permiten máscaras de color. Cuando sea verdadero, el enmascarador permitirá que las máscaras usen cualquier color, no solo blanco, negro o gris. Allow Colored Masks Profile entry in Masker - Allow Colored Masks + Permitir máscaras de colores Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker - Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). + Umbral máximo para que un píxel se considere blanquecino. El color medio a lo largo del borde de una máscara puede no ser blanco puro, por lo que para evitar ligeros tonos blanquecinos sobre un fondo blanco puro, cualquier valor más claro que este umbral se redondeará a blanco puro. Rango de valores: negro (0) a blanco puro (255). @@ -3601,13 +3601,13 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser - Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + Número máximo de hilos a utilizar para la eliminación de ruido. Puede dejarlo sin especificar para utilizar todos los subprocesos disponibles. Reduzca este valor si tiene problemas de memoria, que aparecerán como bloqueos aleatorios. Max Threads Profile entry in Denoiser - Max Threads + Máximo de hilos @@ -3712,73 +3712,73 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter - Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. + Desvanezca los bordes de la máscara de cobertura pintada en esta cantidad de píxeles para mezclar suavemente las partes pintadas con el resto de la imagen.<br> Si ve contornos tenues después de pintar, aumente el valor de pintado mínimo y establezca éste a la mitad de esa cantidad. Inpainting Fade Radius Profile entry in Inpainter - Inpainting Fade Radius + Radio de desvanecimiento Detect per box Language option for OCR - Detect per box + Detectar por caja Detect per page Language option for OCR - Detect per page + Detectar por página Japanese Language option for OCR - Japanese + Japonés English Language option for OCR - English + Inglés Korean Language option for OCR - Korean + Coreano Korean (vertical) Language option for OCR - Korean (vertical) + Coreano (vertical) Chinese - Simplified Language option for OCR - Chinese - Simplified + Chino simplificado Chinese - Traditional Language option for OCR - Chinese - Traditional + Chino tradicional Albanian Language option for OCR - Albanian + Albano Arabic Language option for OCR - Arabic + Arabic @@ -3796,13 +3796,13 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Bengali Language option for OCR - Bengali + Bengalí Bulgarian Language option for OCR - Bulgarian + Búlgaro @@ -3814,37 +3814,37 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Catalan; Valencian Language option for OCR - Catalan; Valencian + Catalán; Valenciano Croatian Language option for OCR - Croatian + Croata Czech Language option for OCR - Czech + Checo Danish Language option for OCR - Danish + Danés Dutch; Flemish Language option for OCR - Dutch; Flemish + Neerlandés; Flamenco Esperanto Language option for OCR - Esperanto + Esperanto @@ -3857,61 +3857,61 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Finnish Filipino Language option for OCR - Finnish + Finlandés French Language option for OCR - French + Francés Georgian Language option for OCR - Georgian + Georgiano German Language option for OCR - German + Alemán Greek Language option for OCR - Greek + Griego Hebrew Language option for OCR - Hebrew + Hebreo Hindi Language option for OCR - Hindi + Hindú Hungarian Language option for OCR - Hungarian + Húngaro Indonesian Language option for OCR - Indonesian + Indonesio Italian Language option for OCR - Italian + Italiano @@ -3923,109 +3923,109 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Latin Language option for OCR - Latin + Latín Lithuanian Language option for OCR - Lithuanian + Lituano Luxembourgish Language option for OCR - Luxembourgish + Luxemburgués Malay Language option for OCR - Malay + Malayo Mongolian Language option for OCR - Mongolian + Mongol Nepali Language option for OCR - Nepali + Nepalí Norwegian Language option for OCR - Norwegian + Noruego Persian Language option for OCR - Persian + Persa Polish Language option for OCR - Polish + Polaco Portuguese Language option for OCR - Portuguese + Portugués Romanian; Moldavian Language option for OCR - Romanian; Moldavian + Rumano; Moldavo Russian Language option for OCR - Russian + Ruso Serbian Language option for OCR - Serbian + Serbio Serbian - Latin Language option for OCR - Serbian - Latin + Serbio - Latín Slovak Language option for OCR - Slovak + Eslovaco Slovenian Language option for OCR - Slovenian + Eslovaco Spanish; Castilian Language option for OCR - Spanish; Castilian + Español, Castellano Swedish Language option for OCR - Swedish + Sueco @@ -4049,25 +4049,25 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Thai Language option for OCR - Thai + Tailandés Turkish Language option for OCR - Turkish + Turco Ukrainian Language option for OCR - Ukrainian + Ucraniano Vietnamese Language option for OCR - Vietnamese + Vietnamita @@ -4238,7 +4238,7 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin px Pixel unit - px + px @@ -4250,7 +4250,7 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin All Cores The limit on the number of CPU cores to use, or here, no limit. - All Cores + Todos los Núcleos @@ -4262,37 +4262,37 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin auto OCR Engine option - auto + automático manga-ocr OCR Engine option - manga-ocr + manga-ocr tesseract OCR Engine option - tesseract + tesseract auto Reading order option - auto + automático manga Reading order option - manga + manga comic Reading order option - comic + comic @@ -4304,73 +4304,73 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin PSD Bulk Layered export option - PSD Bulk + PSD por lotes PSD Per Image Layered export option - PSD Per Image + PSD por imagen Detect per box Language code option - Detect per box + Detectar por caja Detect per page Language code option - Detect per page + Detectar por página Japanese Language code option - Japanese + Japonés English Language code option - English + Inglés Korean Language code option - Korean + Coreano Korean (vertical) Language code option - Korean (vertical) + Coreano (vertical) Chinese - Simplified Language code option - Chinese - Simplified + Chino simplificado Chinese - Traditional Language code option - Chinese - Traditional + Chino tradicional Albanian Language code option - Albanian + Albanés Arabic Language code option - Arabic + Árabe @@ -4388,121 +4388,121 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Bengali Language code option - Bengali + Bengalí Bulgarian Language code option - Bulgarian + Búlgaro Burmese Language code option - Burmese + Birmano Catalan; Valencian Language code option - Catalan; Valencian + Catalán; Valenciano Croatian Language code option - Croatian + Croata Czech Language code option - Czech + Checo Danish Language code option - Danish + Danés Dutch; Flemish Language code option - Dutch; Flemish + Neerlandés; Flamenco Esperanto Language code option - Esperanto + Esperanto Estonian Language code option - Estonian + Estonio Finnish Language code option - Finnish + Finlandés French Language code option - French + Francés Georgian Language code option - Georgian + Gregoriano German Language code option - German + Alemán Greek Language code option - Greek + Griego Hebrew Language code option - Hebrew + Hebreo Hindi Language code option - Hindi + Hindú Hungarian Language code option - Hungarian + Húngaro Indonesian Language code option - Indonesian + Indonesio Italian Language code option - Italian + Italiano @@ -4514,61 +4514,61 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Latin Language code option - Latin + Latín Lithuanian Language code option - Lithuanian + Lituano Luxembourgish Language code option - Luxembourgish + Luxemburgués Malay Language code option - Malay + Malayo Mongolian Language code option - Mongolian + Mongol Nepali Language code option - Nepali + Nepalí Norwegian Language code option - Norwegian + Noruego Persian Language code option - Persian + Persa Polish Language code option - Polish + Polaco Portuguese Language code option - Portuguese + Portugués @@ -4580,43 +4580,43 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Russian Language code option - Russian + Ruso Serbian Language code option - Serbian + Serbio Serbian - Latin Language code option - Serbian - Latin + Serbio - Latín Slovak Language code option - Slovak + Eslovaco Slovenian Language code option - Slovenian + Eslovenio Spanish; Castilian Language code option - Spanish; Castilian + Español; Castellano Swedish Language code option - Swedish + Sueco @@ -4640,25 +4640,25 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Thai Language code option - Thai + Tailandés Turkish Language code option - Turkish + Turco Ukrainian Language code option - Ukrainian + Ucraniano Vietnamese Language code option - Vietnamese + Vietnamita @@ -4673,7 +4673,7 @@ Puede descargar el modelo manualmente desde <a href="https://github.com/ The regex pattern "{value}" for {key} is invalid, reverting to default. - The regex pattern "{value}" for {key} is invalid, reverting to default. + El patrón regex "{value}" para {key} no es válido, volviendo al predeterminado. From 4434296d03f0f8b852a3f96a3fea764def7a4caf Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Sun, 4 Jan 2026 14:08:25 +0100 Subject: [PATCH 03/18] New translations panelcleaner_source.ts (Vietnamese) --- translations/PanelCleaner_vi_VN.ts | 4737 ++++++++++++++++++++++++++++ 1 file changed, 4737 insertions(+) create mode 100644 translations/PanelCleaner_vi_VN.ts diff --git a/translations/PanelCleaner_vi_VN.ts b/translations/PanelCleaner_vi_VN.ts new file mode 100644 index 0000000..cab2298 --- /dev/null +++ b/translations/PanelCleaner_vi_VN.ts @@ -0,0 +1,4737 @@ + + + + + + + + default + reserved profile name + default + + + + Profile file {profile_path} already exists. + Profile file {profile_path} already exists. + + + + Profile {profile_name} created. + Profile {profile_name} created. + + + + Profile name cannot be empty. + Profile name cannot be empty. + + + + Profile name already in use. + Profile name already in use. + + + + Profile name is reserved. + Profile name is reserved. + + + + + + + Unknown + Unknown + + + + RGB + Color mode + RGB + + + + CMYK + Color mode + CMYK + + + + Grayscale + Color mode + Grayscale + + + + 1-bit + Color mode + 1-bit + + + + Palette + Color mode + Palette + + + + Error downloading file from url: {url} + Error downloading file from url: {url} + + + + Error downloading file from url: {url} +Failed to save the file to {save_path} + Error downloading file from url: {url} +Failed to save the file to {save_path} + + + + Error downloading file from url: {url} +The file content is different from expected. + Error downloading file from url: {url} +The file content is different from expected. + + + + + + + + + + + N/A + N/A + + + + OCR Analytics + OCR Analytics + + + + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) + + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) + + + + + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) + + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) + + + + + +Small box sizes: + + +Small box sizes: + + + + + No not-removed small boxes found. + + No not-removed small boxes found. + + + + + +Removed bubbles: + + +Removed bubbles: + + + + + Page {path}: {text} + + Page {path}: {text} + + + + + Small boxes + Small boxes + + + + Removed boxes + Removed boxes + + + + + + Failed + Failed + + + + Perfect + Perfect + + + + Total + + Total + + + + + Mask Fitment Analytics + Mask Fitment Analytics + + + + Total + Total + + + + Total boxes + Total boxes + + + + Masks succeeded + Masks succeeded + + + + Masks failed + Masks failed + + + + Perfect masks + Perfect masks + + + + Average border deviation + Average border deviation + + + + +Mask usage by thickness (in pixels): + + +Mask usage by thickness (in pixels): + + + + + Mask ({thickness}px) + Mask ({thickness}px) + + + + Box mask + Box mask + + + + All bubbles were successfully masked. + + All bubbles were successfully masked. + + + + + +Pages with failures / total: + + +Pages with failures / total: + + + + + Denoising Analytics + Denoising Analytics + + + + Total masks + Total masks + + + + Masks denoised + Masks denoised + + + + Minimum deviation to denoise + Minimum deviation to denoise + + + + Maximum allowed deviation + Maximum allowed deviation + + + + Standard deviation around masks: + + + Standard deviation around masks: + + + + + + Denoised + Denoised + + + + Inpainting Analytics + Inpainting Analytics + + + + Inpainting performed + Inpainting performed + + + + Average thickness + Average thickness + + + + Minimum thickness + Minimum thickness + + + + Maximum thickness + Maximum thickness + + + + Outline thickness around text inpainted: + + + Outline thickness around text inpainted: + + + + + + Inpainted + Inpainted + + + + File Error + File Error + + + + Failed to open file. + Failed to open file. + + + + Failed to access the file. + Failed to access the file. + + + + An error occurred. + An error occurred. + + + + The CSV file format requires exactly 6 columns. + The CSV file format requires exactly 6 columns. + + + + The coordinates must be integers. + The coordinates must be integers. + + + + The box coordinates may not exceed 2 billion. + The box coordinates may not exceed 2 billion. + + + + The file format was not recognized. + The file format was not recognized. + + + + The CSV file must start with a header row, followed by data rows. + The CSV file must start with a header row, followed by data rows. + + + + The file path is missing. + The file path is missing. + + + + The file path is invalid. + The file path is invalid. + + + + OCR Parse Errors + OCR Parse Errors + + + + Multiple analytics match the image: {path} + Multiple analytics match the image: {path} + + + + Multiple images match the file path: {path} + Multiple images match the file path: {path} + + + + Error, no version found. + Error, no version found. + + + + CUDA Error + CUDA Error + + + + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> + + + + Inpainting model not found. + Inpainting model not found. + + + + + filename + filename + + + + + startx + startx + + + + + starty + starty + + + + + endx + endx + + + + + endy + endy + + + + + text + text + + + + + Saved detected text to {output_file} + Saved detected text to {output_file} + + + + + Failed to write detected text to {output_file} + Failed to write detected text to {output_file} + + + + + Save Failed + Save Failed + + + + + Failed to write detected text to file. + Failed to write detected text to file. + + + + Inpainter (experimental) + Inpainter (experimental) + + + + Invalid type for img_paths: {paths} + Invalid type for img_paths: {paths} + + + + Unsupported image format: {suffix} for {path} + Unsupported image format: {suffix} for {path} + + + + Image path {img_path} does not exist. + Image path {img_path} does not exist. + + + + Clean mask + layered export + + Clean mask + + + + + Denoised mask + layered export + + Denoised mask + + + + + Inpainting mask + layered export + + Inpainting mask + + + + + Isolated text + layered export + + Isolated text + + + + + Page + layered export + + Page + + + + + %id + %id + + + + %i + %i + + + + %od + %od + + + + %o + %o + + + + %p + %p + + + + About + + + About Panel Cleaner + About Panel Cleaner + + + + # Panel Cleaner + +An AI-powered tool to clean manga panels. + # Panel Cleaner + +An AI-powered tool to clean manga panels. + + + + Version: + Version: + + + + Author: + Author: + + + + Toolkit: + Toolkit: + + + + License: + License: + + + + Source code: + Source code: + + + + CTextEdit + + + Clear + Clear + + + + ColorButton + + + Select Color + Select Color + + + + ErrorDialog + + + Copy to Clipboard + Copy to Clipboard + + + + Open Issue Tracker + Open Issue Tracker + + + + Terminate Panel Cleaner + Terminate Panel Cleaner + + + + Close + Close + + + + Note: Name "{name}" was hidden + Note: Name "{name}" was hidden + + + + Failed to load log. + Failed to load log. + + + + FileManagerExtension + + + Unsupported system + Unsupported system + + + + This system is not supported. + This system is not supported. + + + + Install for Dolphin + Install for Dolphin + + + + Uninstall for Dolphin + Uninstall for Dolphin + + + + The extension will be installed at {path}. This doesn't require root permissions. + The extension will be installed at {path}. This doesn't require root permissions. + + + + The extension file will be deleted. This doesn't require root permissions. + The extension file will be deleted. This doesn't require root permissions. + + + + Install for Windows Explorer + Install for Windows Explorer + + + + Uninstall for Windows Explorer + Uninstall for Windows Explorer + + + + The extension will be installed in the Windows registry.This requires admin permissions. + The extension will be installed in the Windows registry.This requires admin permissions. + + + + The extension will be uninstalled from the Windows registry.This requires admin permissions. + The extension will be uninstalled from the Windows registry.This requires admin permissions. + + + + Dolphin extension installed successfully. + Dolphin extension installed successfully. + + + + Installation successful + Installation successful + + + + Dolphin extension installed successfully. +You may need to log out and log back in (or restart) for the changes to take effect. + Dolphin extension installed successfully. +You may need to log out and log back in (or restart) for the changes to take effect. + + + + + Installation failed + Installation failed + + + + Failed to install Dolphin extension. + Failed to install Dolphin extension. + + + + Failed to install Windows Explorer extension. + Failed to install Windows Explorer extension. + + + + + Uninstallation successful + Uninstallation successful + + + + Dolphin extension uninstalled successfully. + Dolphin extension uninstalled successfully. + + + + Dolphin extension was not installed. + Dolphin extension was not installed. + + + + + Uninstallation failed + Uninstallation failed + + + + Failed to uninstall Dolphin extension. + Failed to uninstall Dolphin extension. + + + + Failed to uninstall Windows Explorer extension. + Failed to uninstall Windows Explorer extension. + + + + File Manager Extension + File Manager Extension + + + + ### Add or Remove Integration with your File Browser + +This feature adds a menu option to your file browser's context menu (which appears when you right-click) for folders or image files. + +It allows you to use Panel Cleaner directly from the file browser, eliminating the need to open the application separately. + +By default, this uses the built-in profile. However, you can change this default setting by choosing **Profile** and then **Set Default** from the application's menu bar. + ### Add or Remove Integration with your File Browser + +This feature adds a menu option to your file browser's context menu (which appears when you right-click) for folders or image files. + +It allows you to use Panel Cleaner directly from the file browser, eliminating the need to open the application separately. + +By default, this uses the built-in profile. However, you can change this default setting by choosing **Profile** and then **Set Default** from the application's menu bar. + + + + FileTable + + + Original size in pixels (width × height) + Original size in pixels (width × height) + + + + Processing size in pixels (width × height), scale factor + Processing size in pixels (width × height), scale factor + + + + + Loading Failed + Loading Failed + + + + Failed to discover images: {path} + Failed to discover images: {path} + + + + Failed to load images: {path} + Failed to load images: {path} + + + + Unsupported TIFF files + Unsupported TIFF files + + + + The following 5-channel TIFF files are not supported: + + The following 5-channel TIFF files are not supported: + + + + + Duplicate file + Duplicate file + + + + File "{path}" is already in the table. + File "{path}" is already in the table. + + + + Ignore All + Ignore All + + + + split + split + + + + Number of boxes removed by the OCR model / total boxes + Number of boxes removed by the OCR model / total boxes + + + + Number of boxes that failed to generate a mask / total boxes + Number of boxes that failed to generate a mask / total boxes + + + + Number of boxes that were perfectly masked / total boxes + Number of boxes that were perfectly masked / total boxes + + + + Number of boxes that were denoised / total boxes + Number of boxes that were denoised / total boxes + + + + Number of boxes that were inpainted + Number of boxes that were inpainted + + + + Image not loaded + Image not loaded + + + + Please wait until the image has loaded. + Please wait until the image has loaded. + + + + Open individual cleaning + Open individual cleaning + + + + Remove from list + Remove from list + + + + Remove all files from list + Remove all files from list + + + + Failed to load image + Failed to load image + + + + Failed to load image {file_path}. + Failed to load image {file_path}. + + + + Failed to dispatch image. + Failed to dispatch image. + + + + Dispatch Failed + Dispatch Failed + + + + Select files + Select files + + + + Images + Images + + + + Select directory + Select directory + + + + ImageDetails + + + Zoom in + Zoom in + + + + Zoom out + Zoom out + + + + Reset zoom + Reset zoom + + + + Zoom to fit + Zoom to fit + + + + Refresh + Refresh + + + + Generating... + Generating... + + + + Position: + Position: + + + + Size: + Size: + + + + ImageDetailsWidget + + + Copy Image to Clipboard + Copy Image to Clipboard + + + + + Export Image + Export Image + + + + OCR Image + OCR Image + + + + Loading Error + Loading Error + + + + Failed to load image '{path}' + Failed to load image '{path}' + + + + Generate Me + Generate Me + + + + + Impossible Request + Impossible Request + + + + Denoising is disabled in the current profile, this output can't be generated. +Please enable denoising in the profile settings and try again. + Denoising is disabled in the current profile, this output can't be generated. +Please enable denoising in the profile settings and try again. + + + + Inpainting is disabled in the current profile, this output can't be generated. +Please enable inpainting in the profile settings and try again. + Inpainting is disabled in the current profile, this output can't be generated. +Please enable inpainting in the profile settings and try again. + + + + Image not found. + Image not found. + + + + Image at {path} does not exist: + Image at {path} does not exist: + + + + Export failed + Export failed + + + + Failed to export image: + Failed to export image: + + + + Output Failed + Output Failed + + + + Output generation failed: + Output generation failed: + + + + Profile check failed + Profile check failed + + + + Profile change check failed: + Profile change check failed: + + + + ImageMatchOverview + + + image + image + + + + images + images + + + + result + result + + + + results + results + + + + Proceed With Selection + Proceed With Selection + + + + new image + new image + + + + new images + new images + + + + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. + + + + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. + + + + Failed to load image + Failed to load image + + + + Failed to load image {file_path}. + Failed to load image {file_path}. + + + + Dispatch Failed + Dispatch Failed + + + + Failed to dispatch image. + Failed to dispatch image. + + + + Image Selection + Image Selection + + + + Deselect All + Deselect All + + + + Select All + Select All + + + + Unselect images without matching OCR data + Unselect images without matching OCR data + + + + Image File + Image File + + + + OCR Data + Can also call this statistics. + OCR Data + + + + IssueReporter + + + Report an Issue + Report an Issue + + + + Consider including the log for the session that had an issue, if applicable. + Consider including the log for the session that had an issue, if applicable. + + + + View Logs: + View Logs: + + + + Log file is at: + Log file is at: + + + + Copy to Clipboard + Copy to Clipboard + + + + Open Issue Tracker + Open Issue Tracker + + + + Close + Close + + + + Note: Name "{name}" was hidden + Note: Name "{name}" was hidden + + + + Corrupted log session + Corrupted log session + + + + Current session + Current session + + + + Today + Today + + + + Error + Error + + + + Errors + Errors + + + + Critical + Critical + + + + Criticals + Criticals + + + + License + + + License Agreement + License Agreement + + + + MainWindow + + + Open Config + Statusbar button + Open Config + + + + Open Log + Statusbar button + Open Log + + + + CUDA Enabled + CUDA Enabled + + + + Select Output Directory + Select Output Directory + + + + Aborting... + Aborting... + + + + Loading OCR model... + Loading OCR model... + + + + Loaded OCR model. + Loaded OCR model. + + + + Error + Error + + + + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. + +Continue anyway? + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. + +Continue anyway? + + + + Delete Models + Delete Models + + + + Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. + Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. + + + + Encountered error: + Encountered error: + + + + MPS Enabled + MPS Enabled + + + + + Inpainting Model Update + Inpainting Model Update + + + + A new version of the inpainting model is available. +You can delete the model later if you don't want to upgrade yet. +Switch to the new model? + A new version of the inpainting model is available. +You can delete the model later if you don't want to upgrade yet. +Switch to the new model? + + + + Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. + Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. + + + + Layout Reset + Layout Reset + + + + The window layout has been reset. The default layout will be used next time. + The window layout has been reset. The default layout will be used next time. + + + + Multiple Instances + Multiple Instances + + + + Found process ID in lock file: + Found process ID in lock file: + + + + Failed to delete models. + Failed to delete models. + + + + Delete Failed + Delete Failed + + + + Models Deleted + Models Deleted + + + + The models were deleted. Would you like to download them again? + The models were deleted. Would you like to download them again? + + + + Not Supported + Not Supported + + + + Unsupported system or package format + Unsupported system or package format + + + + Shutdown + Shutdown + + + + After processing, the system will shut down. + After processing, the system will shut down. + + + + You have review options enabled, these will need to be manually closed before the action can start. + You have review options enabled, these will need to be manually closed before the action can start. + + + + You have disabled writing output. + You have disabled writing output. + + + + Shutting down system... + Shutting down system... + + + + Running post action "{action}"... + Running post action "{action}"... + + + + Restart Required + Restart Required + + + + The language has been changed. Please restart the application for the changes to take effect. + The language has been changed. Please restart the application for the changes to take effect. + + + + + System Language + System Language + + + + Import Profile + Import Profile + + + + Profile Files (*.conf) + Profile Files (*.conf) + + + + Profile Imported + Profile Imported + + + + Import Error + Import Error + + + + Failed to Delete + Failed to Delete + + + + The default profile cannot be deleted. + The default profile cannot be deleted. + + + + Delete Profile + Delete Profile + + + + Are you sure you want to delete the profile {profile_name}? + Are you sure you want to delete the profile {profile_name}? + + + + Delete Error + Delete Error + + + + Failed to delete the profile. + Failed to delete the profile. + + + + The profile '{profile}' has unsaved changes. +Switching profiles will discard changes to the current profile. + The profile '{profile}' has unsaved changes. +Switching profiles will discard changes to the current profile. + + + + Unsaved changes + Unsaved changes + + + + + Load Error + Load Error + + + + + Failed to load profile: + Failed to load profile: + + + + Tesseract OCR is not installed or not found + Tesseract OCR is not installed or not found + + + + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. +Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. +Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> + + + + Unsupported Language + Unsupported Language + + + + The language '{lang_name}' is not supported by any of your current OCR engines. +Check the online documentation for how to add support for more languages. + The language '{lang_name}' is not supported by any of your current OCR engines. +Check the online documentation for how to add support for more languages. + + + + Profile Created + Profile Created + + + + Create Error + Create Error + + + + Failed to save profile to {profile_path} + Failed to save profile to {profile_path} + + + + + Save Error + Save Error + + + + Failed to save profile. + Failed to save profile. + + + + Profile saved to {profile_path} + Profile saved to {profile_path} + + + + Failed to save config. + Failed to save config. + + + + Failed to save the new profile to the configuration file. + Failed to save the new profile to the configuration file. + + + + + No Files + No Files + + + + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. + + + + No Outputs + No Outputs + + + + No outputs were requested. Please select at least one output before cleaning. + No outputs were requested. Please select at least one output before cleaning. + + + + + cleaned + cleaned + + + + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. + + + + Open OCR Output File + Open OCR Output File + + + + OCR Output Files (*.txt *.csv) + OCR Output Files (*.txt *.csv) + + + + Parse Error + Parse Error + + + + Failed to parse the OCR output file. + Failed to parse the OCR output file. + + + + File Exists + File Exists + + + + The file '{output_path}' already exists. Overwrite? + The file '{output_path}' already exists. Overwrite? + + + + Export Images + Export Images + + + + Would you like to export the cleaned images? + Would you like to export the cleaned images? + + + + Export OCR Results + Export OCR Results + + + + After processing, the "{action}" action will be executed. + After processing, the "{action}" action will be executed. + + + + + Post action canceled. + Post action canceled. + + + + Un-Applied Changes + Un-Applied Changes + + + + You have un-applied changes to the profile. Continue anyway? + You have un-applied changes to the profile. Continue anyway? + + + + Would you like to export the OCR results? + Would you like to export the OCR results? + + + + Processing Finished + Processing Finished + + + + Finished processing all files. + Finished processing all files. + + + + Processing Aborted + Processing Aborted + + + + Processing aborted. + Processing aborted. + + + + Processing Error + Processing Error + + + + Encountered an error while processing files. + Encountered an error while processing files. + + + + Panel Cleaner + Panel Cleaner + + + + + Profile + Profile + + + + Default + Default + + + + Apply + Apply + + + + + Save + Save + + + + Reset All + Reset All + + + + Drag and Drop Images or Folders Here + Drag and Drop Images or Folders Here + + + + Images + Images + + + + + File + File + + + + Size + Size + + + + Processing Size + The size that the picture is shrunk down to for processing. + Processing Size + + + + File Size + File Size + + + + Color Mode + Like RGB, grayscale etc. + Color Mode + + + + Analytics + Can also call this statistics. + Analytics + + + + Process + Process + + + + Cleaning + Cleaning + + + + OCR + OCR + + + + Edit Existing Output + Edit Existing Output + + + + Abort + Abort + + + + Start + Start + + + + Output + Output + + + + Cleaned Image + Cleaned Image + + + + Mask + Mask + + + + Isolated Text + Isolated Text + + + + + Review Output + Review Output + + + + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> + + + + Save Output + Save Output + + + + Output Directory: + Output Directory: + + + + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. + The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. + + + + Plain Text + Plain Text + + + + When checked, a review window will open upon process completion. + When checked, a review window will open upon process completion. + + + + Output File: + Output File: + + + + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> + + + + detected_text.txt + The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. + detected_text.txt + + + + Cancel Action + Cancel Action + + + + Close + Close + + + + Generating: + This is present progressive, as in "[currently] Generating:" + Generating: + + + + Current Step: + Current Step: + + + + + %v / %m + %v / %m + + + + Images Processed: + As in "images processed: 42" with progress bar. + Images Processed: + + + + Set Default + Set Default + + + + Settings + Settings + + + + Theme + As in color theme + Theme + + + + Language + Language + + + + Help + Help + + + + Add Files... + Add Files... + + + + Remove All Files + Remove All Files + + + + Remove File + Remove File + + + + Help Translate Panel Cleaner + Help Translate Panel Cleaner + + + + Report an Issue... + Report an Issue... + + + + Simulate Exception + Simulate Exception + + + + Add File Browser Extension... + Add File Browser Extension... + + + + Show Supported OCR Languages + Show Supported OCR Languages + + + + + Reset Window Layout Preferences + Reset Window Layout Preferences + + + + Show Out Of Memory Warnings + Show Out Of Memory Warnings + + + + Post-Run Actions + Post-Run Actions + + + + Configure Post-Run Actions... + Configure Post-Run Actions... + + + + New + New + + + + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> + + + + CSV Spreadsheet + CSV Spreadsheet + + + + Delete + Delete + + + + Import... + Import... + + + + Online Documentation + Online Documentation + + + + About Panel Cleaner + About Panel Cleaner + + + + View License + View License + + + + Save as... + Save as... + + + + System + System + + + + Dark + Dark + + + + Light + Light + + + + Add Folder... + Add Folder... + + + + Delete Machine Learning Models + Delete Machine Learning Models + + + + Download Machine Learning Models + Download Machine Learning Models + + + + Donate + Donate + + + + MemoryWatcher + + + + RAM usage has reached {mem}% + RAM usage has reached {mem}% + + + + ModelDownloader + + + Downloading Machine Learning Models + Downloading Machine Learning Models + + + + Model data required to run Panel Cleaner: + Model data required to run Panel Cleaner: + + + + OCR model: + OCR model: + + + + + + Already downloaded + Already downloaded + + + + Text Detector model (CUDA) + Text Detector model (CUDA) + + + + Text Detector model (CPU) + Text Detector model (CPU) + + + + + + + + Download Failed + Download Failed + + + + OCR download errors + OCR download errors + + + + ETA + estimated time of completion + ETA + + + + + Inpainting model + Inpainting model + + + + NewProfile + + + New Profile + New Profile + + + + The default profile cannot be overwritten. Save a new profile with your changes: + The default profile cannot be overwritten. Save a new profile with your changes: + + + + Use the default location + Use the default location + + + + Use a different location: + Use a different location: + + + + Profile Name: + Profile Name: + + + + NewProfileDialog + + + Please enter a name. + Please enter a name. + + + + This name is used by a builtin profile. + This name is used by a builtin profile. + + + + A profile with this name already exists. + A profile with this name already exists. + + + + The selected directory does not exist. + The selected directory does not exist. + + + + OCRLanguageSupport + + + OCR Language Support + OCR Language Support + + + + Code + Code + + + + Language + Can also call this statistics. + Language + + + + MangaOCR + MangaOCR + + + + Tesseract + Tesseract + + + + How to install Tesseract + How to install Tesseract + + + + How to install Tesseract language packs + How to install Tesseract language packs + + + + OcrReview + + + Review + Review + + + + Icon Size: + Icon Size: + + + + Previous image + Previous image + + + + Next image + Next image + + + + Zoom in + Zoom in + + + + Zoom out + Zoom out + + + + Reset zoom + Reset zoom + + + + Zoom to fit + Zoom to fit + + + + With Boxes + With Boxes + + + + Original + Original + + + + Move box order up + Move box order up + + + + Move box order down + Move box order down + + + + Add a new box + Add a new box + + + + Delete current box + Delete current box + + + + Recover this deleted box + Recover this deleted box + + + + Reset changes to this image + Reset changes to this image + + + + Reset All + Reset All + + + + OCR new boxes: + OCR new boxes: + + + + Done + Done + + + + Box + Box + + + + Text + Text + + + + OcrReviewWindow + + + No OCR + No OCR + + + + OCR Error + OCR Error + + + + Encountered error: + Encountered error: + + + + Reset Bubbles + Reset Bubbles + + + + Are you sure you want to reset all boxes for this image? + Are you sure you want to reset all boxes for this image? + + + + New + New + + + + OCR Warning + OCR Warning + + + + The bubble is very large. Are you sure you want to OCR it? + The bubble is very large. Are you sure you want to OCR it? + + + + Finish Review + Finish Review + + + + Are you sure you want to finish the review? + Are you sure you want to finish the review? + + + + image + image + + + + images + images + + + + Loading Error + Loading Error + + + + Failed to load image '{path}' + Failed to load image '{path}' + + + + OutputReview + + + Review + Review + + + + Icon Size: + Icon Size: + + + + Previous image + Previous image + + + + Next image + Next image + + + + Zoom in + Zoom in + + + + Zoom out + Zoom out + + + + Reset zoom + Reset zoom + + + + Zoom to fit + Zoom to fit + + + + View Mode: + View Mode: + + + + Side-by-Side + Side-by-Side + + + + Swipe + Swipe + + + + Onion + Onion + + + + Difference + Difference + + + + Overlay + Overlay + + + + Done + Done + + + + OutputReviewWindow + + + Finish Review + Finish Review + + + + Are you sure you want to finish the review? + Are you sure you want to finish the review? + + + + image + image + + + + images + images + + + + Loading Error + Loading Error + + + + Failed to load image '{path}' + Failed to load image '{path}' + + + + PostActionConfiguration + + + Configure Post-Run Actions + Configure Post-Run Actions + + + + Always perform the action after processing + Always perform the action after processing + + + + Wait before performing an action: + Wait before performing an action: + + + + s + unit of time: seconds + s + + + + <html><head/><body><p>If you set a wait time greater than 0, a window will open upon process completion. This window will show a countdown and allow you to preview as well as edit the command.</p></body></html> + <html><head/><body><p>If you set a wait time greater than 0, a window will open upon process completion. This window will show a countdown and allow you to preview as well as edit the command.</p></body></html> + + + + Power Off Command: + Power Off Command: + + + + shutdown + shutdown + + + + Custom Commands: + Custom Commands: + + + + Move command up + Move command up + + + + Move command down + Move command down + + + + New + New + + + + Delete + Delete + + + + Action Name + Action Name + + + + Command or Script File Path + Command or Script File Path + + + + Cancel the custom action if processing failed + Cancel the custom action if processing failed + + + + <html><head/><body><p>The Power Off action will run regardless of this setting.</p></body></html> + <html><head/><body><p>The Power Off action will run regardless of this setting.</p></body></html> + + + + Placeholder Tokens: + Placeholder Tokens: + + + + Input Files + Input Files + + + + %i + Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. + %i + + + + <html><head/><body><p>The individual file paths of each file processed, as a space separated list, with quotation marks where necessary.</p></body></html> + <html><head/><body><p>The individual file paths of each file processed, as a space separated list, with quotation marks where necessary.</p></body></html> + + + + Input Directory + Input Directory + + + + %id + Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. + %id + + + + <html><head/><body><p>Group files by parent directory and list these, as a space separated list, with quotation marks where necessary.</p></body></html> + <html><head/><body><p>Group files by parent directory and list these, as a space separated list, with quotation marks where necessary.</p></body></html> + + + + Output Files + Output Files + + + + %o + Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. + %o + + + + <html><head/><body><p>The individual file paths of each file created as an output, as a space separated list, with quotation marks where necessary.</p></body></html> + <html><head/><body><p>The individual file paths of each file created as an output, as a space separated list, with quotation marks where necessary.</p></body></html> + + + + Output Directory + Output Directory + + + + %od + Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. + %od + + + + <html><head/><body><p>Group output files by parent directory and list these, as a space separated list, with quotation marks where necessary.</p></body></html> + <html><head/><body><p>Group output files by parent directory and list these, as a space separated list, with quotation marks where necessary.</p></body></html> + + + + Profile Used + Profile Used + + + + %p + Yes, you may change this, so long as all tokens are different and the directory version is longer than the file version. So, a and ab are allowed for file and directory, but not the other way around. + %p + + + + <html><head/><body><p>The name of the profile used, with quotation marks if necessary.</p></body></html> + <html><head/><body><p>The name of the profile used, with quotation marks if necessary.</p></body></html> + + + + Action names cannot be empty. + Action names cannot be empty. + + + + Action names must be unique. + Action names must be unique. + + + + The name "{shutdown_command}" is reserved. + The name "{shutdown_command}" is reserved. + + + + The Flatpak sandbox will likely prevent actions from working. + The Flatpak sandbox will likely prevent actions from working. + + + + PostActionRunner + + + Post-Action Runner + Post-Action Runner + + + + An error occurred during the process. + An error occurred during the process. + + + + Full Command: + Full Command: + + + + Run Again + Run Again + + + + Skip Countdown + Skip Countdown + + + + Resume Countdown + Resume Countdown + + + + Pause Countdown + Pause Countdown + + + + Close + Close + + + + Cancel Action + Cancel Action + + + + Cancel Shutdown + Cancel Shutdown + + + + Shutting down the system + Shutting down the system + + + + Running the action "{action_name}" + Running the action "{action_name}" + + + + Remaining time: {time} seconds + Remaining time: {time} seconds + + + + Remaining time: 1 second + Remaining time: 1 second + + + + Process Steps + + + The original image with scaling applied (if needed). + Step description in the image details view, step: Input, output: Input + The original image with scaling applied (if needed). + + + + Input + Step name in the image details view + Input + + + + Input + Output name in the image details view + Input + + + + The rough mask generated by the AI. + Step description in the image details view, step: Text Detection, output: AI Mask + The rough mask generated by the AI. + + + + Text Detection + Step name in the image details view + Text Detection + + + + AI Mask + Output name in the image details view + AI Mask + + + + The unfiltered box data generated by the AI. + Step description in the image details view, step: Text Detection, output: Raw Boxes + The unfiltered box data generated by the AI. + + + + Raw Boxes + Output name in the image details view + Raw Boxes + + + + The outlines of the text boxes the AI found. + Step description in the image details view, step: Preprocessor, output: Initial Boxes + The outlines of the text boxes the AI found. + + + + Preprocessor + Step name in the image details view + Preprocessor + + + + Initial Boxes + Output name in the image details view + Initial Boxes + + + + The final boxes after expanding, merging and filtering unneeded boxes with OCR. +Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. + Step description in the image details view, step: Preprocessor, output: Final Boxes + The final boxes after expanding, merging and filtering unneeded boxes with OCR. +Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. + + + + Final Boxes + Output name in the image details view + Final Boxes + + + + The mask of the merged boxes. + Step description in the image details view, step: Masker, output: Box Mask + The mask of the merged boxes. + + + + Masker + Step name in the image details view + Masker + + + + Box Mask + Output name in the image details view + Box Mask + + + + The rough text detection mask with everything outside the box mask cut out. + Step description in the image details view, step: Masker, output: Cut Mask + The rough text detection mask with everything outside the box mask cut out. + + + + Cut Mask + Output name in the image details view + Cut Mask + + + + The different steps of growth around the cut mask displayed in different colors. + Step description in the image details view, step: Masker, output: Mask Layers + The different steps of growth around the cut mask displayed in different colors. + + + + Mask Layers + Output name in the image details view + Mask Layers + + + + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). + Step description in the image details view, step: Masker, output: Fitment Quality + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). + + + + Fitment Quality + Output name in the image details view + Fitment Quality + + + + The input image with the final mask overlaid in color. + Step description in the image details view, step: Masker, output: Mask Overlay + The input image with the final mask overlaid in color. + + + + Mask Overlay + Output name in the image details view + Mask Overlay + + + + The collection of masks for each bubble that fit best. + Step description in the image details view, step: Masker, output: Final Mask + The collection of masks for each bubble that fit best. + + + + Final Mask + Output name in the image details view + Final Mask + + + + The text layer isolated from the input image. + Step description in the image details view, step: Masker, output: Isolated Text + The text layer isolated from the input image. + + + + Isolated Text + Output name in the image details view + Isolated Text + + + + The input image with the final mask applied. + Step description in the image details view, step: Masker, output: Masked Output + The input image with the final mask applied. + + + + Masked Output + Output name in the image details view + Masked Output + + + + The masks that required denoising, to be overlaid on the final mask when exporting. + Step description in the image details view, step: Denoiser, output: Denoise Mask + The masks that required denoising, to be overlaid on the final mask when exporting. + + + + Denoiser + Step name in the image details view + Denoiser + + + + Denoise Mask + Output name in the image details view + Denoise Mask + + + + The input image with the denoised mask applied. + Step description in the image details view, step: Denoiser, output: Denoised Output + The input image with the denoised mask applied. + + + + Denoised Output + Output name in the image details view + Denoised Output + + + + The inpainted sections around the text that was poorly cleaned, if at all. + Step description in the image details view, step: Inpainter, output: Inpainted Mask + The inpainted sections around the text that was poorly cleaned, if at all. + + + + Inpainter + Step name in the image details view + Inpainter + + + + Inpainted Mask + Output name in the image details view + Inpainted Mask + + + + The input image: cleaned, denoised (if enabled), and inpainted. + Step description in the image details view, step: Inpainter, output: Inpainted Output + The input image: cleaned, denoised (if enabled), and inpainted. + + + + Inpainted Output + Output name in the image details view + Inpainted Output + + + + <This is just a placeholder, translating this string is not necessary> + Step description in the image details view, step: None, output: Write Output + <This is just a placeholder, translating this string is not necessary> + + + + Write Output + Output name in the image details view + Write Output + + + + Profile + + + General + Profile section title + General + + + + About this profile: + Profile comment in General + About this profile: + + + + Notes + Profile entry in General + Notes + + + + Preferred file type to save the cleaned image as. + Profile comment in General + Preferred file type to save the cleaned image as. + + + + Preferred File Type + Profile entry in General + Preferred File Type + + + + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. + Profile comment in General + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. + + + + Preferred Mask File Type + Profile entry in General + Preferred Mask File Type + + + + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. + Profile comment in General + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. + + + + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. + Profile comment in General + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. + + + + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. + Profile comment in General + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. + + + + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. + Profile comment in General + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. + + + + Input Height Lower Target + Profile entry in General + Input Height Lower Target + + + + Input Height Upper Target + Profile entry in General + Input Height Upper Target + + + + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). + Profile comment in General + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). + + + + Split Long Strips + Profile entry in General + Split Long Strips + + + + Preferred height to split long strips at. + Profile comment in General + Preferred height to split long strips at. + + + + Preferred Split Height + Profile entry in General + Preferred Split Height + + + + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. + Profile comment in General + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. + + + + Split Tolerance Margin + Profile entry in General + Split Tolerance Margin + + + + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. + Profile comment in General + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. + + + + Long Strip Aspect Ratio + Profile entry in General + Long Strip Aspect Ratio + + + + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. + Profile comment in General + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. + + + + Merge After Split + Profile entry in General + Merge After Split + + + + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + Profile comment in General + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + + + + Max Threads Export + Profile entry in General + Max Threads Export + + + + Text Detector + Profile section title + Text Detector + + + + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> + Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> + Profile comment in Text Detector + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> + + + + Model Path + Profile entry in Text Detector + Model Path + + + + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! + Profile comment in Text Detector + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! + + + + Concurrent Models + Profile entry in Text Detector + Concurrent Models + + + + Preprocessor + Profile section title + Preprocessor + + + + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. + Profile comment in Preprocessor + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. + + + + Minimum size of any box to keep it. + Profile comment in Preprocessor + Minimum size of any box to keep it. + + + + Box Min Size + Profile entry in Preprocessor + Box Min Size + + + + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. + Profile comment in Preprocessor + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. + + + + Suspicious Box Min Size + Profile entry in Preprocessor + Suspicious Box Min Size + + + + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. + Profile comment in Preprocessor + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. + + + + Box Overlap Threshold + Profile entry in Preprocessor + Box Overlap Threshold + + + + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. + Profile comment in Preprocessor + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. + + + + OCR Enabled + Profile entry in Preprocessor + OCR Enabled + + + + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. + Profile comment in Preprocessor + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. + + + + OCR Use Tesseract + Profile entry in Preprocessor + OCR Use Tesseract + + + + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. + Profile comment in Preprocessor + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. + + + + OCR Language + Profile entry in Preprocessor + OCR Language + + + + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. + Profile comment in Preprocessor + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. + + + + OCR Engine + Profile entry in Preprocessor + OCR Engine + + + + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. + Profile comment in Preprocessor + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. + + + + Reading Order + Profile entry in Preprocessor + Reading Order + + + + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. + Profile comment in Preprocessor + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. + + + + OCR Max Size + Profile entry in Preprocessor + OCR Max Size + + + + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. + Profile comment in Preprocessor + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. + + + + OCR Blacklist Pattern + Profile entry in Preprocessor + OCR Blacklist Pattern + + + + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. + The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. + Profile comment in Preprocessor + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. + + + + OCR Strict Language + Profile entry in Preprocessor + OCR Strict Language + + + + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. + Profile comment in Preprocessor + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. + + + + Box Padding Initial + Profile entry in Preprocessor + Box Padding Initial + + + + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. + Profile comment in Preprocessor + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. + + + + Box Right Padding Initial + Profile entry in Preprocessor + Box Right Padding Initial + + + + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. + Profile comment in Preprocessor + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. + + + + Box Padding Extended + Profile entry in Preprocessor + Box Padding Extended + + + + Box Right Padding Extended + Profile entry in Preprocessor + Box Right Padding Extended + + + + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. + Profile comment in Preprocessor + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. + + + + Box Reference Padding + Profile entry in Preprocessor + Box Reference Padding + + + + Masker + Profile section title + Masker + + + + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + Profile comment in Masker + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + + + + Max Threads + Profile entry in Masker + Max Threads + + + + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. + Profile comment in Masker + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. + + + + Mask Growth Step Pixels + Profile entry in Masker + Mask Growth Step Pixels + + + + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. + Profile comment in Masker + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. + + + + Mask Growth Steps + Profile entry in Masker + Mask Growth Steps + + + + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. + Profile comment in Masker + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. + + + + Min Mask Thickness + Profile entry in Masker + Min Mask Thickness + + + + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. + Profile comment in Inpainter + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. + + + + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. + Profile comment in General + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. + + + + Layered Export + Profile entry in General + Layered Export + + + + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. + Profile comment in Masker + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. + + + + Allow Colored Masks + Profile entry in Masker + Allow Colored Masks + + + + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). + Profile comment in Masker + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). + + + + Off White Max Threshold + Profile entry in Masker + Off White Max Threshold + + + + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. + Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. + Profile comment in Masker + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. + + + + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. + Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. + Profile comment in Masker + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. + + + + Mask Improvement Threshold + Profile entry in Masker + Mask Improvement Threshold + + + + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. + Profile comment in Masker + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. + + + + Mask Selection Fast + Profile entry in Masker + Mask Selection Fast + + + + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + Profile comment in Denoiser + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. + + + + Max Threads + Profile entry in Denoiser + Max Threads + + + + Inpainter + Profile section title + Inpainter + + + + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. + Profile comment in Inpainter + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. + + + + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. + Profile comment in Inpainter + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. + + + + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. + Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. + Profile comment in Inpainter + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. + + + + Inpainting Enabled + Profile entry in Inpainter + Inpainting Enabled + + + + Inpainting Min Std Dev + Profile entry in Inpainter + Inpainting Min Std Dev + + + + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. + Profile comment in Inpainter + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. + + + + Inpainting Max Mask Radius + Profile entry in Inpainter + Inpainting Max Mask Radius + + + + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. + Profile comment in Inpainter + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. + + + + Min Inpainting Radius + Profile entry in Inpainter + Min Inpainting Radius + + + + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation + Profile comment in Inpainter + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation + + + + Inpainting Radius Multiplier + Profile entry in Inpainter + Inpainting Radius Multiplier + + + + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. + Profile comment in Inpainter + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. + + + + Max Inpainting Radius + Profile entry in Inpainter + Max Inpainting Radius + + + + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. + After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. + Profile comment in Inpainter + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. + + + + Inpainting Isolation Radius + Profile entry in Inpainter + Inpainting Isolation Radius + + + + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. + Profile comment in Inpainter + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. + + + + Inpainting Fade Radius + Profile entry in Inpainter + Inpainting Fade Radius + + + + Detect per box + Language option for OCR + Detect per box + + + + Detect per page + Language option for OCR + Detect per page + + + + Japanese + Language option for OCR + Japanese + + + + English + Language option for OCR + English + + + + Korean + Language option for OCR + Korean + + + + Korean (vertical) + Language option for OCR + Korean (vertical) + + + + Chinese - Simplified + Language option for OCR + Chinese - Simplified + + + + Chinese - Traditional + Language option for OCR + Chinese - Traditional + + + + Albanian + Language option for OCR + Albanian + + + + Arabic + Language option for OCR + Arabic + + + + Azerbaijani + Language option for OCR + Azerbaijani + + + + Azerbaijani - Cyrilic + Language option for OCR + Azerbaijani - Cyrilic + + + + Bengali + Language option for OCR + Bengali + + + + Bulgarian + Language option for OCR + Bulgarian + + + + Burmese + Language option for OCR + Burmese + + + + Catalan; Valencian + Language option for OCR + Catalan; Valencian + + + + Croatian + Language option for OCR + Croatian + + + + Czech + Language option for OCR + Czech + + + + Danish + Language option for OCR + Danish + + + + Dutch; Flemish + Language option for OCR + Dutch; Flemish + + + + Esperanto + Language option for OCR + Esperanto + + + + Estonian + Language option for OCR + Estonian + + + + Finnish + Filipino + Language option for OCR + Finnish + + + + French + Language option for OCR + French + + + + Georgian + Language option for OCR + Georgian + + + + German + Language option for OCR + German + + + + Greek + Language option for OCR + Greek + + + + Hebrew + Language option for OCR + Hebrew + + + + Hindi + Language option for OCR + Hindi + + + + Hungarian + Language option for OCR + Hungarian + + + + Indonesian + Language option for OCR + Indonesian + + + + Italian + Language option for OCR + Italian + + + + Kazakh + Language option for OCR + Kazakh + + + + Latin + Language option for OCR + Latin + + + + Lithuanian + Language option for OCR + Lithuanian + + + + Luxembourgish + Language option for OCR + Luxembourgish + + + + Malay + Language option for OCR + Malay + + + + Mongolian + Language option for OCR + Mongolian + + + + Nepali + Language option for OCR + Nepali + + + + Norwegian + Language option for OCR + Norwegian + + + + Persian + Language option for OCR + Persian + + + + Polish + Language option for OCR + Polish + + + + Portuguese + Language option for OCR + Portuguese + + + + Romanian; Moldavian + Language option for OCR + Romanian; Moldavian + + + + Russian + Language option for OCR + Russian + + + + Serbian + Language option for OCR + Serbian + + + + Serbian - Latin + Language option for OCR + Serbian - Latin + + + + Slovak + Language option for OCR + Slovak + + + + Slovenian + Language option for OCR + Slovenian + + + + Spanish; Castilian + Language option for OCR + Spanish; Castilian + + + + Swedish + Language option for OCR + Swedish + + + + Tagalog + Language option for OCR + Tagalog + + + + Tamil + Language option for OCR + Tamil + + + + Telugu + Language option for OCR + Telugu + + + + Thai + Language option for OCR + Thai + + + + Turkish + Language option for OCR + Turkish + + + + Ukrainian + Language option for OCR + Ukrainian + + + + Vietnamese + Language option for OCR + Vietnamese + + + + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. + Profile comment in Masker + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. + + + + Mask Max Standard Deviation + Profile entry in Masker + Mask Max Standard Deviation + + + + Color to use for the debug mask. + Profile comment in Masker + Color to use for the debug mask. + + + + Debug Mask Color + Profile entry in Masker + Debug Mask Color + + + + Denoiser + Profile section title + Denoiser + + + + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. + Profile comment in Denoiser + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. + + + + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. + Profile comment in Denoiser + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. + + + + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. + Profile comment in Denoiser + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. + + + + Denoising Enabled + Profile entry in Denoiser + Denoising Enabled + + + + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. + Profile comment in Denoiser + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. + + + + Noise Min Standard Deviation + Profile entry in Denoiser + Noise Min Standard Deviation + + + + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. + Profile comment in Denoiser + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. + + + + The thickness of an outline to denoise around a mask. + Profile comment in Denoiser + The thickness of an outline to denoise around a mask. + + + + Noise Outline Size + Profile entry in Denoiser + Noise Outline Size + + + + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. + Profile comment in Denoiser + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. + + + + Noise Fade Radius + Profile entry in Denoiser + Noise Fade Radius + + + + Set to True to support colored images. + Profile comment in Denoiser + Set to True to support colored images. + + + + Colored Images + Profile entry in Denoiser + Colored Images + + + + How strongly to denoise the image. Higher values mean more denoising. + Profile comment in Denoiser + How strongly to denoise the image. Higher values mean more denoising. + + + + Filter Strength + Profile entry in Denoiser + Filter Strength + + + + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. + Profile comment in Denoiser + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. + + + + Color Filter Strength + Profile entry in Denoiser + Color Filter Strength + + + + Size in pixels of the template patch that is used to compute weights. Should be odd. + Profile comment in Denoiser + Size in pixels of the template patch that is used to compute weights. Should be odd. + + + + Template Window Size + Profile entry in Denoiser + Template Window Size + + + + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. + Profile comment in Denoiser + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. + + + + Search Window Size + Profile entry in Denoiser + Search Window Size + + + + ProfileOptionWidget + + + Reset to default + Generic reset button tooltip + Reset to default + + + + px + Pixel unit + px + + + + px² + Pixel squared unit + px² + + + + All Cores + The limit on the number of CPU cores to use, or here, no limit. + All Cores + + + + Same as image + Profile default option for the file type + Same as image + + + + auto + OCR Engine option + auto + + + + manga-ocr + OCR Engine option + manga-ocr + + + + tesseract + OCR Engine option + tesseract + + + + auto + Reading order option + auto + + + + manga + Reading order option + manga + + + + comic + Reading order option + comic + + + + None + Layered export option + None + + + + PSD Bulk + Layered export option + PSD Bulk + + + + PSD Per Image + Layered export option + PSD Per Image + + + + Detect per box + Language code option + Detect per box + + + + Detect per page + Language code option + Detect per page + + + + Japanese + Language code option + Japanese + + + + English + Language code option + English + + + + Korean + Language code option + Korean + + + + Korean (vertical) + Language code option + Korean (vertical) + + + + Chinese - Simplified + Language code option + Chinese - Simplified + + + + Chinese - Traditional + Language code option + Chinese - Traditional + + + + Albanian + Language code option + Albanian + + + + Arabic + Language code option + Arabic + + + + Azerbaijani + Language code option + Azerbaijani + + + + Azerbaijani - Cyrilic + Language code option + Azerbaijani - Cyrilic + + + + Bengali + Language code option + Bengali + + + + Bulgarian + Language code option + Bulgarian + + + + Burmese + Language code option + Burmese + + + + Catalan; Valencian + Language code option + Catalan; Valencian + + + + Croatian + Language code option + Croatian + + + + Czech + Language code option + Czech + + + + Danish + Language code option + Danish + + + + Dutch; Flemish + Language code option + Dutch; Flemish + + + + Esperanto + Language code option + Esperanto + + + + Estonian + Language code option + Estonian + + + + Finnish + Language code option + Finnish + + + + French + Language code option + French + + + + Georgian + Language code option + Georgian + + + + German + Language code option + German + + + + Greek + Language code option + Greek + + + + Hebrew + Language code option + Hebrew + + + + Hindi + Language code option + Hindi + + + + Hungarian + Language code option + Hungarian + + + + Indonesian + Language code option + Indonesian + + + + Italian + Language code option + Italian + + + + Kazakh + Language code option + Kazakh + + + + Latin + Language code option + Latin + + + + Lithuanian + Language code option + Lithuanian + + + + Luxembourgish + Language code option + Luxembourgish + + + + Malay + Language code option + Malay + + + + Mongolian + Language code option + Mongolian + + + + Nepali + Language code option + Nepali + + + + Norwegian + Language code option + Norwegian + + + + Persian + Language code option + Persian + + + + Polish + Language code option + Polish + + + + Portuguese + Language code option + Portuguese + + + + Romanian; Moldavian + Language code option + Romanian; Moldavian + + + + Russian + Language code option + Russian + + + + Serbian + Language code option + Serbian + + + + Serbian - Latin + Language code option + Serbian - Latin + + + + Slovak + Language code option + Slovak + + + + Slovenian + Language code option + Slovenian + + + + Spanish; Castilian + Language code option + Spanish; Castilian + + + + Swedish + Language code option + Swedish + + + + Tagalog + Language code option + Tagalog + + + + Tamil + Language code option + Tamil + + + + Telugu + Language code option + Telugu + + + + Thai + Language code option + Thai + + + + Turkish + Language code option + Turkish + + + + Ukrainian + Language code option + Ukrainian + + + + Vietnamese + Language code option + Vietnamese + + + + ProfileToolBox + + + <html>The Text Detector model path {value} does not exist, reverting to default. +You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> + <html>The Text Detector model path {value} does not exist, reverting to default. +You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> + + + + The regex pattern "{value}" for {key} is invalid, reverting to default. + The regex pattern "{value}" for {key} is invalid, reverting to default. + + + + RegexLineEdit + + + Regex is valid + Regex is valid + + + + SetupGreeter + + + Downloading Machine Learning Models + Downloading Machine Learning Models + + + + ## Setting up Panel Cleaner + ## Setting up Panel Cleaner + + + + Before Panel Cleaner can begin, it requires the following machine learning models. Press **Automatically Download** to allow Panel Cleaner to download these files to get started (Internet connection required, total space approx. 700 MiB) + Before Panel Cleaner can begin, it requires the following machine learning models. Press **Automatically Download** to allow Panel Cleaner to download these files to get started (Internet connection required, total space approx. 700 MiB) + + + + Text Detection: + Text Detection: + + + + + + Saved in: + Saved in: + + + + Optical Character Recognition (OCR): + Optical Character Recognition (OCR): + + + + Inpainting: + Inpainting: + + + + Automatically Download + Automatically Download + + + + Abort + Abort + + + + (only the comictextdetector file is needed) + (only the comictextdetector file is needed) + + + From 81835f966dce5e0e50b2a2d0484aab16c23f202d Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:06 +0100 Subject: [PATCH 04/18] New translations panelcleaner_source.ts (Vietnamese) --- translations/PanelCleaner_vi_VN.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_vi_VN.ts b/translations/PanelCleaner_vi_VN.ts index cab2298..6eeddf6 100644 --- a/translations/PanelCleaner_vi_VN.ts +++ b/translations/PanelCleaner_vi_VN.ts @@ -4,7 +4,7 @@ - + default reserved profile name default @@ -20,54 +20,54 @@ Profile {profile_name} created. - + Profile name cannot be empty. Profile name cannot be empty. - + Profile name already in use. Profile name already in use. - + Profile name is reserved. Profile name is reserved. - - - - + + + + Unknown Unknown - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Grayscale - + 1-bit Color mode 1-bit - + Palette Color mode Palette @@ -78,45 +78,45 @@ Error downloading file from url: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Error downloading file from url: {url} Failed to save the file to {save_path} - + Error downloading file from url: {url} The file content is different from expected. Error downloading file from url: {url} The file content is different from expected. + - - - + + + - - - + + N/A N/A - + OCR Analytics OCR Analytics - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) @@ -155,66 +155,66 @@ Removed bubbles: - + Small boxes Small boxes - + Removed boxes Removed boxes - - + + Failed Failed - + Perfect Perfect - + Total Total - + Mask Fitment Analytics Mask Fitment Analytics - + Total Total - + Total boxes Total boxes - + Masks succeeded Masks succeeded - + Masks failed Masks failed - + Perfect masks Perfect masks - + Average border deviation Average border deviation @@ -233,7 +233,7 @@ Mask usage by thickness (in pixels): Mask ({thickness}px) - + Box mask Box mask @@ -254,27 +254,27 @@ Pages with failures / total: - + Denoising Analytics Denoising Analytics - + Total masks Total masks - + Masks denoised Masks denoised - + Minimum deviation to denoise Minimum deviation to denoise - + Maximum allowed deviation Maximum allowed deviation @@ -293,27 +293,27 @@ Pages with failures / total: Denoised - + Inpainting Analytics Inpainting Analytics - + Inpainting performed Inpainting performed - + Average thickness Average thickness - + Minimum thickness Minimum thickness - + Maximum thickness Maximum thickness @@ -372,7 +372,7 @@ Pages with failures / total: The file format was not recognized. - + The CSV file must start with a header row, followed by data rows. The CSV file must start with a header row, followed by data rows. @@ -402,7 +402,7 @@ Pages with failures / total: Multiple images match the file path: {path} - + Error, no version found. Error, no version found. @@ -412,7 +412,7 @@ Pages with failures / total: CUDA Error - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pages with failures / total: Inpainting model not found. - + filename filename - + startx startx - + starty starty - + endx endx - + endy endy - + text text - + Saved detected text to {output_file} Saved detected text to {output_file} - + Failed to write detected text to {output_file} Failed to write detected text to {output_file} - + Save Failed Save Failed - + Failed to write detected text to file. Failed to write detected text to file. @@ -534,7 +534,7 @@ Pages with failures / total: - + Page layered export @@ -620,7 +620,7 @@ An AI-powered tool to clean manga panels. ColorButton - + Select Color Select Color @@ -666,7 +666,7 @@ An AI-powered tool to clean manga panels. Unsupported system - + This system is not supported. This system is not supported. @@ -681,12 +681,12 @@ An AI-powered tool to clean manga panels. Uninstall for Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. The extension will be installed at {path}. This doesn't require root permissions. - + The extension file will be deleted. This doesn't require root permissions. The extension file will be deleted. This doesn't require root permissions. @@ -701,17 +701,17 @@ An AI-powered tool to clean manga panels. Uninstall for Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. The extension will be installed in the Windows registry.This requires admin permissions. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. The extension will be uninstalled from the Windows registry.This requires admin permissions. - + Dolphin extension installed successfully. Dolphin extension installed successfully. @@ -721,7 +721,7 @@ An AI-powered tool to clean manga panels. Installation successful - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Dolphin extension installed successfully. @@ -801,12 +801,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) Original size in pixels (width × height) - + Processing size in pixels (width × height), scale factor Processing size in pixels (width × height), scale factor @@ -832,7 +832,7 @@ By default, this uses the built-in profile. However, you can change this default Unsupported TIFF files - + The following 5-channel TIFF files are not supported: The following 5-channel TIFF files are not supported: @@ -859,27 +859,27 @@ By default, this uses the built-in profile. However, you can change this default split - + Number of boxes removed by the OCR model / total boxes Number of boxes removed by the OCR model / total boxes - + Number of boxes that failed to generate a mask / total boxes Number of boxes that failed to generate a mask / total boxes - + Number of boxes that were perfectly masked / total boxes Number of boxes that were perfectly masked / total boxes - + Number of boxes that were denoised / total boxes Number of boxes that were denoised / total boxes - + Number of boxes that were inpainted Number of boxes that were inpainted @@ -1027,14 +1027,14 @@ By default, this uses the built-in profile. However, you can change this default Impossible Request - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting is disabled in the current profile, this output can't be generated. @@ -1056,7 +1056,7 @@ Please enable inpainting in the profile settings and try again. Export failed - + Failed to export image: Failed to export image: @@ -1099,7 +1099,7 @@ Please enable inpainting in the profile settings and try again. result - + results results @@ -1114,17 +1114,17 @@ Please enable inpainting in the profile settings and try again. new image - + new images new images - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. @@ -1223,12 +1223,12 @@ Please enable inpainting in the profile settings and try again. Note: Name "{name}" was hidden - + Corrupted log session Corrupted log session - + Current session Current session @@ -1243,7 +1243,7 @@ Please enable inpainting in the profile settings and try again. Error - + Errors Errors @@ -1253,7 +1253,7 @@ Please enable inpainting in the profile settings and try again. Critical - + Criticals Criticals @@ -1269,49 +1269,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button Open Config - + Open Log Statusbar button Open Log - + CUDA Enabled CUDA Enabled - + Select Output Directory Select Output Directory - + Aborting... Aborting... - + Loading OCR model... Loading OCR model... - + Loaded OCR model. Loaded OCR model. - + Error Error - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1320,7 +1320,7 @@ Continue anyway? Continue anyway? - + Delete Models Delete Models @@ -1330,23 +1330,23 @@ Continue anyway? Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. - + Encountered error: Encountered error: - + MPS Enabled MPS Enabled - - + + Inpainting Model Update Inpainting Model Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1360,148 +1360,148 @@ Switch to the new model? Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. - + Layout Reset Layout Reset - + The window layout has been reset. The default layout will be used next time. The window layout has been reset. The default layout will be used next time. - + Multiple Instances Multiple Instances - + Found process ID in lock file: Found process ID in lock file: - + Failed to delete models. Failed to delete models. - + Delete Failed Delete Failed - + Models Deleted Models Deleted - + The models were deleted. Would you like to download them again? The models were deleted. Would you like to download them again? - + Not Supported Not Supported - + Unsupported system or package format Unsupported system or package format - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Restart Required - + The language has been changed. Please restart the application for the changes to take effect. The language has been changed. Please restart the application for the changes to take effect. - - + + System Language System Language - + Import Profile Import Profile - + Profile Files (*.conf) Profile Files (*.conf) - + Profile Imported Profile Imported - + Import Error Import Error - + Failed to Delete Failed to Delete - + The default profile cannot be deleted. The default profile cannot be deleted. - + Delete Profile Delete Profile - + Are you sure you want to delete the profile {profile_name}? Are you sure you want to delete the profile {profile_name}? - + Delete Error Delete Error - + Failed to delete the profile. Failed to delete the profile. @@ -1513,217 +1513,217 @@ Switching profiles will discard changes to the current profile. Switching profiles will discard changes to the current profile. - + Unsaved changes Unsaved changes - - + + Load Error Load Error - - + + Failed to load profile: Failed to load profile: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Profile Created - + Create Error Create Error - + Failed to save profile to {profile_path} Failed to save profile to {profile_path} - - + + Save Error Save Error - + Failed to save profile. Failed to save profile. - + Profile saved to {profile_path} Profile saved to {profile_path} - + Failed to save config. Failed to save config. - + Failed to save the new profile to the configuration file. Failed to save the new profile to the configuration file. - - + + No Files No Files - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. - + No Outputs No Outputs - + No outputs were requested. Please select at least one output before cleaning. No outputs were requested. Please select at least one output before cleaning. - - + + cleaned cleaned - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists File Exists - + The file '{output_path}' already exists. Overwrite? The file '{output_path}' already exists. Overwrite? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Processing Finished - + Finished processing all files. Finished processing all files. - + Processing Aborted Processing Aborted - + Processing aborted. Processing aborted. - + Processing Error Processing Error - + Encountered an error while processing files. Encountered an error while processing files. @@ -1734,7 +1734,7 @@ Check the online documentation for how to add support for more languages. - + Profile Profile @@ -1749,360 +1749,365 @@ Check the online documentation for how to add support for more languages.Apply - - + + Save Save - + Reset All Reset All - + Drag and Drop Images or Folders Here Drag and Drop Images or Folders Here - + Images Images - - + + File File - + Size Size - + Processing Size The size that the picture is shrunk down to for processing. Processing Size - + File Size File Size - + Color Mode Like RGB, grayscale etc. Color Mode - + Analytics Can also call this statistics. Analytics - + Process Process - + Cleaning Cleaning - + OCR OCR - + Edit Existing Output Edit Existing Output - + Abort Abort - + Start Start - + Output Output - + Cleaned Image Cleaned Image - + Mask Mask - + Isolated Text Isolated Text - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Save Output - + Output Directory: Output Directory: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. You can use a relative path to create a subfolder at the image's original location, or use an absolute path. - + Plain Text Plain Text - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Output File: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Generating: - + Current Step: Current Step: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Images Processed: - + Set Default Set Default - + Settings Settings - + Theme As in color theme Theme - + Language Language - + Help Help - + Add Files... Add Files... - + Remove All Files Remove All Files - + Remove File Remove File - + Help Translate Panel Cleaner Help Translate Panel Cleaner - + Report an Issue... Report an Issue... - + Simulate Exception Simulate Exception - + Add File Browser Extension... Add File Browser Extension... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New New - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Delete - + Import... Import... - + Online Documentation Online Documentation - + About Panel Cleaner About Panel Cleaner - + View License View License - + Save as... Save as... - + System System - + Dark Dark - + Light Light - + Add Folder... Add Folder... - + Delete Machine Learning Models Delete Machine Learning Models - + Download Machine Learning Models Download Machine Learning Models - + Donate Donate @@ -2134,19 +2139,19 @@ Check the online documentation for how to add support for more languages.OCR model: - + Already downloaded Already downloaded - + Text Detector model (CUDA) Text Detector model (CUDA) - + Text Detector model (CPU) Text Detector model (CPU) @@ -2165,7 +2170,7 @@ Check the online documentation for how to add support for more languages.OCR download errors - + ETA estimated time of completion ETA @@ -2743,7 +2748,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2824,73 +2829,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input The original image with scaling applied (if needed). - + Input Step name in the image details view Input - + Input Output name in the image details view Input - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask The rough mask generated by the AI. - + Text Detection Step name in the image details view Text Detection - + AI Mask Output name in the image details view AI Mask - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes The unfiltered box data generated by the AI. - + Raw Boxes Output name in the image details view Raw Boxes - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes The outlines of the text boxes the AI found. - + Preprocessor Step name in the image details view Preprocessor - + Initial Boxes Output name in the image details view Initial Boxes - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2898,181 +2903,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. - + Final Boxes Output name in the image details view Final Boxes - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask The mask of the merged boxes. - + Masker Step name in the image details view Masker - + Box Mask Output name in the image details view Box Mask - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask The rough text detection mask with everything outside the box mask cut out. - + Cut Mask Output name in the image details view Cut Mask - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers The different steps of growth around the cut mask displayed in different colors. - + Mask Layers Output name in the image details view Mask Layers - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). - + Fitment Quality Output name in the image details view Fitment Quality - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay The input image with the final mask overlaid in color. - + Mask Overlay Output name in the image details view Mask Overlay - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask The collection of masks for each bubble that fit best. - + Final Mask Output name in the image details view Final Mask - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text The text layer isolated from the input image. - + Isolated Text Output name in the image details view Isolated Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output The input image with the final mask applied. - + Masked Output Output name in the image details view Masked Output - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask The masks that required denoising, to be overlaid on the final mask when exporting. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Denoise Mask - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output The input image with the denoised mask applied. - + Denoised Output Output name in the image details view Denoised Output - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask The inpainted sections around the text that was poorly cleaned, if at all. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Inpainted Mask - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output The input image: cleaned, denoised (if enabled), and inpainted. - + Inpainted Output Output name in the image details view Inpainted Output - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3081,1141 +3086,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title General - + About this profile: Profile comment in General About this profile: - + Notes Profile entry in General Notes - + Preferred file type to save the cleaned image as. Profile comment in General Preferred file type to save the cleaned image as. - + Preferred File Type Profile entry in General Preferred File Type - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Preferred file type to save the mask as. Only image formats that allow for transparency are supported. - + Preferred Mask File Type Profile entry in General Preferred Mask File Type - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. - + Input Height Lower Target Profile entry in General Input Height Lower Target - + Input Height Upper Target Profile entry in General Input Height Upper Target - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Text Detector - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> - + Model Path Profile entry in Text Detector Model Path - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! - + Concurrent Models Profile entry in Text Detector Concurrent Models - + Preprocessor Profile section title Preprocessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Minimum size of any box to keep it. - + Box Min Size Profile entry in Preprocessor Box Min Size - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. - + Suspicious Box Min Size Profile entry in Preprocessor Suspicious Box Min Size - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. - + Box Overlap Threshold Profile entry in Preprocessor Box Overlap Threshold - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. - + OCR Enabled Profile entry in Preprocessor OCR Enabled - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. - + OCR Use Tesseract Profile entry in Preprocessor OCR Use Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. - + OCR Engine Profile entry in Preprocessor OCR Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. - + Reading Order Profile entry in Preprocessor Reading Order - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. - + OCR Max Size Profile entry in Preprocessor OCR Max Size - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR Blacklist Pattern - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor OCR Strict Language - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. - + Box Padding Initial Profile entry in Preprocessor Box Padding Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. - + Box Right Padding Initial Profile entry in Preprocessor Box Right Padding Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Box Right Padding Extended - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. - + Box Reference Padding Profile entry in Preprocessor Box Reference Padding - + Masker Profile section title Masker - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. - + Mask Growth Step Pixels Profile entry in Masker Mask Growth Step Pixels - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. - + Mask Growth Steps Profile entry in Masker Mask Growth Steps - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. - + Min Mask Thickness Profile entry in Masker Min Mask Thickness - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Off White Max Threshold - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. - + Mask Improvement Threshold Profile entry in Masker Mask Improvement Threshold - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. - + Mask Selection Fast Profile entry in Masker Mask Selection Fast - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting Enabled - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. - + Min Inpainting Radius Profile entry in Inpainter Min Inpainting Radius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation - + Inpainting Radius Multiplier Profile entry in Inpainter Inpainting Radius Multiplier - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. - + Max Inpainting Radius Profile entry in Inpainter Max Inpainting Radius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. - + Inpainting Isolation Radius Profile entry in Inpainter Inpainting Isolation Radius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. - + Mask Max Standard Deviation Profile entry in Masker Mask Max Standard Deviation - + Color to use for the debug mask. Profile comment in Masker Color to use for the debug mask. - + Debug Mask Color Profile entry in Masker Debug Mask Color - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. - + Denoising Enabled Profile entry in Denoiser Denoising Enabled - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. - + Noise Min Standard Deviation Profile entry in Denoiser Noise Min Standard Deviation - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser The thickness of an outline to denoise around a mask. - + Noise Outline Size Profile entry in Denoiser Noise Outline Size - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. - + Noise Fade Radius Profile entry in Denoiser Noise Fade Radius - + Set to True to support colored images. Profile comment in Denoiser Set to True to support colored images. - + Colored Images Profile entry in Denoiser Colored Images - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the image. Higher values mean more denoising. - + Filter Strength Profile entry in Denoiser Filter Strength - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. - + Color Filter Strength Profile entry in Denoiser Color Filter Strength - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Size in pixels of the template patch that is used to compute weights. Should be odd. - + Template Window Size Profile entry in Denoiser Template Window Size - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. - + Search Window Size Profile entry in Denoiser Search Window Size @@ -4242,7 +4247,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4659,14 +4664,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From f211c078894303bccad5ca2dbb794e97539d9e85 Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:07 +0100 Subject: [PATCH 05/18] New translations panelcleaner_source.ts (French) --- translations/PanelCleaner_fr_FR.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_fr_FR.ts b/translations/PanelCleaner_fr_FR.ts index 0724ecc..8078b92 100644 --- a/translations/PanelCleaner_fr_FR.ts +++ b/translations/PanelCleaner_fr_FR.ts @@ -4,7 +4,7 @@ - + default reserved profile name default @@ -20,54 +20,54 @@ Profil {profile_name} créé. - + Profile name cannot be empty. Le nom du profil ne peut pas être vide. - + Profile name already in use. Le nom du profil est déjà utilisé. - + Profile name is reserved. Le nom du profil est réservé. - - - - + + + + Unknown Inconnu - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Niveaux de gris - + 1-bit Color mode 1-bit - + Palette Color mode Palette @@ -78,45 +78,45 @@ Erreur lors du téléchargement du fichier à partir de l'URL : {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Erreur lors du téléchargement du fichier à partir de l'URL : {url} Échec de l'enregistrement du fichier dans {save_path} - + Error downloading file from url: {url} The file content is different from expected. Error downloading file from url: {url} The file content is different from expected. + - - - + + + - - - + + N/A N/A - + OCR Analytics OCR Analytics - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) @@ -155,66 +155,66 @@ Removed bubbles: - + Small boxes Small boxes - + Removed boxes Removed boxes - - + + Failed Failed - + Perfect Perfect - + Total Total - + Mask Fitment Analytics Mask Fitment Analytics - + Total Total - + Total boxes Total boxes - + Masks succeeded Masks succeeded - + Masks failed Masks failed - + Perfect masks Perfect masks - + Average border deviation Average border deviation @@ -233,7 +233,7 @@ Mask usage by thickness (in pixels): Mask ({thickness}px) - + Box mask Box mask @@ -254,27 +254,27 @@ Pages with failures / total: - + Denoising Analytics Denoising Analytics - + Total masks Total masks - + Masks denoised Masks denoised - + Minimum deviation to denoise Minimum deviation to denoise - + Maximum allowed deviation Maximum allowed deviation @@ -293,27 +293,27 @@ Pages with failures / total: Denoised - + Inpainting Analytics Inpainting Analytics - + Inpainting performed Inpainting performed - + Average thickness Average thickness - + Minimum thickness Minimum thickness - + Maximum thickness Maximum thickness @@ -372,7 +372,7 @@ Pages with failures / total: The file format was not recognized. - + The CSV file must start with a header row, followed by data rows. The CSV file must start with a header row, followed by data rows. @@ -402,7 +402,7 @@ Pages with failures / total: Multiple images match the file path: {path} - + Error, no version found. Error, no version found. @@ -412,7 +412,7 @@ Pages with failures / total: CUDA Error - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pages with failures / total: Inpainting model not found. - + filename filename - + startx startx - + starty starty - + endx endx - + endy endy - + text text - + Saved detected text to {output_file} Saved detected text to {output_file} - + Failed to write detected text to {output_file} Failed to write detected text to {output_file} - + Save Failed Save Failed - + Failed to write detected text to file. Failed to write detected text to file. @@ -538,7 +538,7 @@ Pages with failures / total: - + Page layered export @@ -625,7 +625,7 @@ An AI-powered tool to clean manga panels. ColorButton - + Select Color Select Color @@ -671,7 +671,7 @@ An AI-powered tool to clean manga panels. Unsupported system - + This system is not supported. This system is not supported. @@ -686,12 +686,12 @@ An AI-powered tool to clean manga panels. Uninstall for Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. The extension will be installed at {path}. This doesn't require root permissions. - + The extension file will be deleted. This doesn't require root permissions. The extension file will be deleted. This doesn't require root permissions. @@ -706,17 +706,17 @@ An AI-powered tool to clean manga panels. Uninstall for Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. The extension will be installed in the Windows registry.This requires admin permissions. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. The extension will be uninstalled from the Windows registry.This requires admin permissions. - + Dolphin extension installed successfully. Dolphin extension installed successfully. @@ -726,7 +726,7 @@ An AI-powered tool to clean manga panels. Installation successful - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Dolphin extension installed successfully. @@ -806,12 +806,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) Original size in pixels (width × height) - + Processing size in pixels (width × height), scale factor Processing size in pixels (width × height), scale factor @@ -837,7 +837,7 @@ By default, this uses the built-in profile. However, you can change this default Unsupported TIFF files - + The following 5-channel TIFF files are not supported: The following 5-channel TIFF files are not supported: @@ -864,27 +864,27 @@ By default, this uses the built-in profile. However, you can change this default split - + Number of boxes removed by the OCR model / total boxes Number of boxes removed by the OCR model / total boxes - + Number of boxes that failed to generate a mask / total boxes Number of boxes that failed to generate a mask / total boxes - + Number of boxes that were perfectly masked / total boxes Number of boxes that were perfectly masked / total boxes - + Number of boxes that were denoised / total boxes Number of boxes that were denoised / total boxes - + Number of boxes that were inpainted Number of boxes that were inpainted @@ -1032,14 +1032,14 @@ By default, this uses the built-in profile. However, you can change this default Impossible Request - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting is disabled in the current profile, this output can't be generated. @@ -1061,7 +1061,7 @@ Please enable inpainting in the profile settings and try again. Export failed - + Failed to export image: Failed to export image: @@ -1104,7 +1104,7 @@ Please enable inpainting in the profile settings and try again. result - + results results @@ -1119,17 +1119,17 @@ Please enable inpainting in the profile settings and try again. new image - + new images new images - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. @@ -1228,12 +1228,12 @@ Please enable inpainting in the profile settings and try again. Note: Name "{name}" was hidden - + Corrupted log session Corrupted log session - + Current session Current session @@ -1248,7 +1248,7 @@ Please enable inpainting in the profile settings and try again. Error - + Errors Errors @@ -1258,7 +1258,7 @@ Please enable inpainting in the profile settings and try again. Critical - + Criticals Criticals @@ -1274,49 +1274,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button Open Config - + Open Log Statusbar button Open Log - + CUDA Enabled CUDA Enabled - + Select Output Directory Select Output Directory - + Aborting... Aborting... - + Loading OCR model... Loading OCR model... - + Loaded OCR model. Loaded OCR model. - + Error Error - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1325,7 +1325,7 @@ Continue anyway? Continue anyway? - + Delete Models Delete Models @@ -1335,23 +1335,23 @@ Continue anyway? Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. - + Encountered error: Encountered error: - + MPS Enabled MPS Enabled - - + + Inpainting Model Update Inpainting Model Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1365,148 +1365,148 @@ Switch to the new model? Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. - + Layout Reset Layout Reset - + The window layout has been reset. The default layout will be used next time. The window layout has been reset. The default layout will be used next time. - + Multiple Instances Multiple Instances - + Found process ID in lock file: Found process ID in lock file: - + Failed to delete models. Failed to delete models. - + Delete Failed Delete Failed - + Models Deleted Models Deleted - + The models were deleted. Would you like to download them again? The models were deleted. Would you like to download them again? - + Not Supported Not Supported - + Unsupported system or package format Unsupported system or package format - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Restart Required - + The language has been changed. Please restart the application for the changes to take effect. The language has been changed. Please restart the application for the changes to take effect. - - + + System Language System Language - + Import Profile Import Profile - + Profile Files (*.conf) Profile Files (*.conf) - + Profile Imported Profile Imported - + Import Error Import Error - + Failed to Delete Failed to Delete - + The default profile cannot be deleted. The default profile cannot be deleted. - + Delete Profile Delete Profile - + Are you sure you want to delete the profile {profile_name}? Are you sure you want to delete the profile {profile_name}? - + Delete Error Delete Error - + Failed to delete the profile. Failed to delete the profile. @@ -1518,217 +1518,217 @@ Switching profiles will discard changes to the current profile. Switching profiles will discard changes to the current profile. - + Unsaved changes Unsaved changes - - + + Load Error Load Error - - + + Failed to load profile: Failed to load profile: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Profile Created - + Create Error Create Error - + Failed to save profile to {profile_path} Failed to save profile to {profile_path} - - + + Save Error Save Error - + Failed to save profile. Failed to save profile. - + Profile saved to {profile_path} Profile saved to {profile_path} - + Failed to save config. Failed to save config. - + Failed to save the new profile to the configuration file. Failed to save the new profile to the configuration file. - - + + No Files No Files - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. - + No Outputs No Outputs - + No outputs were requested. Please select at least one output before cleaning. No outputs were requested. Please select at least one output before cleaning. - - + + cleaned cleaned - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists File Exists - + The file '{output_path}' already exists. Overwrite? The file '{output_path}' already exists. Overwrite? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Processing Finished - + Finished processing all files. Finished processing all files. - + Processing Aborted Processing Aborted - + Processing aborted. Processing aborted. - + Processing Error Processing Error - + Encountered an error while processing files. Encountered an error while processing files. @@ -1739,7 +1739,7 @@ Check the online documentation for how to add support for more languages. - + Profile Profile @@ -1754,360 +1754,365 @@ Check the online documentation for how to add support for more languages.Apply - - + + Save Save - + Reset All Reset All - + Drag and Drop Images or Folders Here Drag and Drop Images or Folders Here - + Images Images - - + + File File - + Size Size - + Processing Size The size that the picture is shrunk down to for processing. Processing Size - + File Size File Size - + Color Mode Like RGB, grayscale etc. Color Mode - + Analytics Can also call this statistics. Analytics - + Process Process - + Cleaning Cleaning - + OCR OCR - + Edit Existing Output Edit Existing Output - + Abort Abort - + Start Start - + Output Output - + Cleaned Image Cleaned Image - + Mask Mask - + Isolated Text Isolated Text - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Save Output - + Output Directory: Output Directory: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. You can use a relative path to create a subfolder at the image's original location, or use an absolute path. - + Plain Text Plain Text - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Output File: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Generating: - + Current Step: Current Step: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Images Processed: - + Set Default Set Default - + Settings Settings - + Theme As in color theme Theme - + Language Language - + Help Help - + Add Files... Add Files... - + Remove All Files Remove All Files - + Remove File Remove File - + Help Translate Panel Cleaner Help Translate Panel Cleaner - + Report an Issue... Report an Issue... - + Simulate Exception Simulate Exception - + Add File Browser Extension... Add File Browser Extension... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New New - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Delete - + Import... Import... - + Online Documentation Online Documentation - + About Panel Cleaner About Panel Cleaner - + View License View License - + Save as... Save as... - + System System - + Dark Dark - + Light Light - + Add Folder... Add Folder... - + Delete Machine Learning Models Delete Machine Learning Models - + Download Machine Learning Models Download Machine Learning Models - + Donate Donate @@ -2139,19 +2144,19 @@ Check the online documentation for how to add support for more languages.OCR model: - + Already downloaded Already downloaded - + Text Detector model (CUDA) Text Detector model (CUDA) - + Text Detector model (CPU) Text Detector model (CPU) @@ -2170,7 +2175,7 @@ Check the online documentation for how to add support for more languages.OCR download errors - + ETA estimated time of completion ETA @@ -2748,7 +2753,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2829,73 +2834,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input The original image with scaling applied (if needed). - + Input Step name in the image details view Input - + Input Output name in the image details view Input - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask The rough mask generated by the AI. - + Text Detection Step name in the image details view Text Detection - + AI Mask Output name in the image details view AI Mask - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes The unfiltered box data generated by the AI. - + Raw Boxes Output name in the image details view Raw Boxes - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes The outlines of the text boxes the AI found. - + Preprocessor Step name in the image details view Preprocessor - + Initial Boxes Output name in the image details view Initial Boxes - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2903,181 +2908,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. - + Final Boxes Output name in the image details view Final Boxes - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask The mask of the merged boxes. - + Masker Step name in the image details view Masker - + Box Mask Output name in the image details view Box Mask - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask The rough text detection mask with everything outside the box mask cut out. - + Cut Mask Output name in the image details view Cut Mask - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers The different steps of growth around the cut mask displayed in different colors. - + Mask Layers Output name in the image details view Mask Layers - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). - + Fitment Quality Output name in the image details view Fitment Quality - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay The input image with the final mask overlaid in color. - + Mask Overlay Output name in the image details view Mask Overlay - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask The collection of masks for each bubble that fit best. - + Final Mask Output name in the image details view Final Mask - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text The text layer isolated from the input image. - + Isolated Text Output name in the image details view Isolated Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output The input image with the final mask applied. - + Masked Output Output name in the image details view Masked Output - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask The masks that required denoising, to be overlaid on the final mask when exporting. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Denoise Mask - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output The input image with the denoised mask applied. - + Denoised Output Output name in the image details view Denoised Output - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask The inpainted sections around the text that was poorly cleaned, if at all. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Inpainted Mask - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output The input image: cleaned, denoised (if enabled), and inpainted. - + Inpainted Output Output name in the image details view Inpainted Output - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3086,1141 +3091,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title General - + About this profile: Profile comment in General About this profile: - + Notes Profile entry in General Notes - + Preferred file type to save the cleaned image as. Profile comment in General Preferred file type to save the cleaned image as. - + Preferred File Type Profile entry in General Preferred File Type - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Preferred file type to save the mask as. Only image formats that allow for transparency are supported. - + Preferred Mask File Type Profile entry in General Preferred Mask File Type - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. - + Input Height Lower Target Profile entry in General Input Height Lower Target - + Input Height Upper Target Profile entry in General Input Height Upper Target - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Text Detector - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> - + Model Path Profile entry in Text Detector Model Path - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! - + Concurrent Models Profile entry in Text Detector Concurrent Models - + Preprocessor Profile section title Preprocessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Minimum size of any box to keep it. - + Box Min Size Profile entry in Preprocessor Box Min Size - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. - + Suspicious Box Min Size Profile entry in Preprocessor Suspicious Box Min Size - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. - + Box Overlap Threshold Profile entry in Preprocessor Box Overlap Threshold - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. - + OCR Enabled Profile entry in Preprocessor OCR Enabled - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. - + OCR Use Tesseract Profile entry in Preprocessor OCR Use Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. - + OCR Engine Profile entry in Preprocessor OCR Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. - + Reading Order Profile entry in Preprocessor Reading Order - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. - + OCR Max Size Profile entry in Preprocessor OCR Max Size - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR Blacklist Pattern - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor OCR Strict Language - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. - + Box Padding Initial Profile entry in Preprocessor Box Padding Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. - + Box Right Padding Initial Profile entry in Preprocessor Box Right Padding Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Box Right Padding Extended - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. - + Box Reference Padding Profile entry in Preprocessor Box Reference Padding - + Masker Profile section title Masker - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. - + Mask Growth Step Pixels Profile entry in Masker Mask Growth Step Pixels - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. - + Mask Growth Steps Profile entry in Masker Mask Growth Steps - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. - + Min Mask Thickness Profile entry in Masker Min Mask Thickness - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Off White Max Threshold - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. - + Mask Improvement Threshold Profile entry in Masker Mask Improvement Threshold - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. - + Mask Selection Fast Profile entry in Masker Mask Selection Fast - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting Enabled - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. - + Min Inpainting Radius Profile entry in Inpainter Min Inpainting Radius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation - + Inpainting Radius Multiplier Profile entry in Inpainter Inpainting Radius Multiplier - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. - + Max Inpainting Radius Profile entry in Inpainter Max Inpainting Radius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. - + Inpainting Isolation Radius Profile entry in Inpainter Inpainting Isolation Radius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. - + Mask Max Standard Deviation Profile entry in Masker Mask Max Standard Deviation - + Color to use for the debug mask. Profile comment in Masker Color to use for the debug mask. - + Debug Mask Color Profile entry in Masker Debug Mask Color - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. - + Denoising Enabled Profile entry in Denoiser Denoising Enabled - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. - + Noise Min Standard Deviation Profile entry in Denoiser Noise Min Standard Deviation - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser The thickness of an outline to denoise around a mask. - + Noise Outline Size Profile entry in Denoiser Noise Outline Size - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. - + Noise Fade Radius Profile entry in Denoiser Noise Fade Radius - + Set to True to support colored images. Profile comment in Denoiser Set to True to support colored images. - + Colored Images Profile entry in Denoiser Colored Images - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the image. Higher values mean more denoising. - + Filter Strength Profile entry in Denoiser Filter Strength - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. - + Color Filter Strength Profile entry in Denoiser Color Filter Strength - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Size in pixels of the template patch that is used to compute weights. Should be odd. - + Template Window Size Profile entry in Denoiser Template Window Size - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. - + Search Window Size Profile entry in Denoiser Search Window Size @@ -4247,7 +4252,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4664,14 +4669,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From c624df2026da5f999d0a60328a91a3df990d9a1d Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:09 +0100 Subject: [PATCH 06/18] New translations panelcleaner_source.ts (Spanish) --- translations/PanelCleaner_es_ES.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_es_ES.ts b/translations/PanelCleaner_es_ES.ts index fa39b8e..0e240e6 100644 --- a/translations/PanelCleaner_es_ES.ts +++ b/translations/PanelCleaner_es_ES.ts @@ -4,7 +4,7 @@ - + default reserved profile name por defecto @@ -20,54 +20,54 @@ Perfil {profile_name} creado. - + Profile name cannot be empty. El nombre del perfil no puede estar vacío. - + Profile name already in use. Nombre de perfil ya utilizado. - + Profile name is reserved. El nombre del perfil está reservado. - - - - + + + + Unknown Desconocido - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Escala de grises - + 1-bit Color mode 1 bit - + Palette Color mode Paleta @@ -78,45 +78,45 @@ Error al descargar el archivo desde la url: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Error al descargar el archivo desde la url: {url} Error al guardar el archivo en {save_path} - + Error downloading file from url: {url} The file content is different from expected. Error al descargar el archivo desde la url: {url} El contenido del archivo es diferente al esperado. + - - - + + + - - - + + N/A N/A - + OCR Analytics Análisis OCR - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Número de cajas: {num_boxes} | Número de cajas pequeñas: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Número de cajas retiradas: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} de cajas pequeñas) @@ -155,66 +155,66 @@ Burbujas eliminadas: - + Small boxes Cajas pequeñas - + Removed boxes Cajas retiradas - - + + Failed Fallido - + Perfect Perfecto - + Total Total - + Mask Fitment Analytics Análisis de ajuste de máscaras - + Total Total - + Total boxes Total de cajas - + Masks succeeded Las máscaras tuvieron éxito - + Masks failed Máscaras fallidas - + Perfect masks Máscaras perfectas - + Average border deviation Desviación fronteriza media @@ -233,7 +233,7 @@ Uso de la máscara por grosor (en píxeles): Máscara ({thickness}px) - + Box mask Máscara de caja @@ -254,27 +254,27 @@ Páginas con fallos / total: - + Denoising Analytics Análisis de eliminación de ruido - + Total masks Máscaras totales - + Masks denoised Máscaras eliminadas - + Minimum deviation to denoise Desviación mínima de eliminación de ruido - + Maximum allowed deviation Desviación máxima permitida @@ -293,27 +293,27 @@ Páginas con fallos / total: Denoise - + Inpainting Analytics Análisis de pintado - + Inpainting performed Repintado realizado - + Average thickness Espesor medio - + Minimum thickness Espesor mínimo - + Maximum thickness Espesor máximo @@ -372,7 +372,7 @@ Páginas con fallos / total: El formato del archivo no pudo reconocerse. - + The CSV file must start with a header row, followed by data rows. El archivo CSV debe comenzar con una fila de cabecera, seguido de la fila de datos. @@ -402,7 +402,7 @@ Páginas con fallos / total: Múltiples imágenes coinciden con la ruta del archivo: {path} - + Error, no version found. Error, versión no encontrada. @@ -412,7 +412,7 @@ Páginas con fallos / total: Error CUDA. - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Tu GPU no soporta las operaciones CUDA requeridas.<br><br>Prueba con desinstalar las versiones actuales de torch y torchvision e instale en su lugar una versión para CPU (o una versión CUDA diferente).<br>Puedes encontrar más instrucciones aquí: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" Para ver las versiones disponibles.<br><br>Tu versión CUDA actual es: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Páginas con fallos / total: No se ha encontrado el modelo de pintura. - + filename nombre de archivo - + startx startx - + starty starty - + endx finx - + endy endy - + text texto - + Saved detected text to {output_file} Texto detectado guardado en {output_file} - + Failed to write detected text to {output_file} Error al escribir el texto detectado en {output_file} - + Save Failed Guardar fallido - + Failed to write detected text to file. Error al escribir el texto detectado en el archivo. @@ -538,7 +538,7 @@ Páginas con fallos / total: - + Page layered export @@ -625,7 +625,7 @@ Una herramienta de IA para limpiar paneles de manga. ColorButton - + Select Color Seleccionar color @@ -671,7 +671,7 @@ Una herramienta de IA para limpiar paneles de manga. Sistema no compatible - + This system is not supported. Este sistema no es compatible. @@ -686,12 +686,12 @@ Una herramienta de IA para limpiar paneles de manga. Desinstalación de Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. La extensión se instalará en {path}. Esto no'requiere permisos de root. - + The extension file will be deleted. This doesn't require root permissions. Se eliminará el archivo de extensión. Esto no'requiere permisos de root. @@ -706,17 +706,17 @@ Una herramienta de IA para limpiar paneles de manga. Desinstalación del Explorador de Windows - + The extension will be installed in the Windows registry.This requires admin permissions. La extensión se instalará en el registro de Windows, lo que requiere permisos de administrador. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. La extensión se desinstalará del registro de Windows, lo que requiere permisos de administrador. - + Dolphin extension installed successfully. La extensión Dolphin se ha instalado correctamente. @@ -726,7 +726,7 @@ Una herramienta de IA para limpiar paneles de manga. Instalación correcta - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. La extensión Dolphin se ha instalado correctamente. @@ -806,12 +806,12 @@ Por defecto, utiliza el perfil incorporado. No obstante, puede cambiar esta conf FileTable - + Original size in pixels (width × height) Tamaño original en píxeles (ancho × alto) - + Processing size in pixels (width × height), scale factor Tamaño de procesamiento en píxeles (ancho × alto), factor de escala @@ -837,7 +837,7 @@ Por defecto, utiliza el perfil incorporado. No obstante, puede cambiar esta conf Archivos TIFF no compatibles - + The following 5-channel TIFF files are not supported: Los siguientes archivos TIFF de 5 canales no son compatibles: @@ -864,27 +864,27 @@ Por defecto, utiliza el perfil incorporado. No obstante, puede cambiar esta conf separar - + Number of boxes removed by the OCR model / total boxes Número de cajas eliminadas por el modelo OCR / total de cajas - + Number of boxes that failed to generate a mask / total boxes Número de casillas en las que no se generó una máscara / total de casillas - + Number of boxes that were perfectly masked / total boxes Número de casillas en las que no se generó una máscara / total de casillas - + Number of boxes that were denoised / total boxes Número de cajas sin procesar / total de cajas - + Number of boxes that were inpainted Número de cajas pintadas @@ -1032,14 +1032,14 @@ Por defecto, utiliza el perfil incorporado. No obstante, puede cambiar esta conf Petición imposible - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. La eliminación de ruido está desactivada en el perfil actual, por lo que no se puede generar esta salida'. Active la eliminación de ruido en la configuración del perfil y vuelva a intentarlo. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting está desactivado en el perfil actual, esta salida no puede ser generada'. @@ -1061,7 +1061,7 @@ Por favor, active la función inpainting en la configuración del perfil e inté Error de exportación - + Failed to export image: Error al exportar la imagen: @@ -1104,7 +1104,7 @@ Por favor, active la función inpainting en la configuración del perfil e inté resultado - + results resultados @@ -1119,17 +1119,17 @@ Por favor, active la función inpainting en la configuración del perfil e inté Nueva imagen - + new images Nuevas imágenes - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Coincidencia {num_images} {plural_images} con los resultados del OCR. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} Se perderá el OCR huérfano. {plural_result} @@ -1228,12 +1228,12 @@ Por favor, active la función inpainting en la configuración del perfil e inté Nota: El nombre "{name}" estaba oculto - + Corrupted log session Sesión de registro dañada - + Current session Sesión en curso @@ -1248,7 +1248,7 @@ Por favor, active la función inpainting en la configuración del perfil e inté Error - + Errors Errores @@ -1258,7 +1258,7 @@ Por favor, active la función inpainting en la configuración del perfil e inté Crítica - + Criticals Críticas @@ -1274,49 +1274,49 @@ Por favor, active la función inpainting en la configuración del perfil e inté MainWindow - + Open Config Statusbar button Abrir configuración - + Open Log Statusbar button Registro abierto - + CUDA Enabled CUDA habilitado - + Select Output Directory Seleccionar directorio de salida - + Aborting... Abortar... - + Loading OCR model... Cargando modelo OCR... - + Loaded OCR model. Modelo OCR cargado. - + Error Error - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1325,7 +1325,7 @@ Continue anyway? ¿Continuar de todos modos? - + Delete Models Borrar modelos @@ -1335,23 +1335,23 @@ Continue anyway? ¿Está seguro de que desea eliminar los modelos de aprendizaje automático? Esto hará que la limpieza y el OCR sean imposibles hasta que se descarguen de nuevo. - + Encountered error: Error encontrado: - + MPS Enabled MPS activado - - + + Inpainting Model Update Actualización del modelo de pintado. - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1365,148 +1365,148 @@ Puede eliminar el modelo más adelante si no desea actualizarlo todavía'. Se conserva el modelo antiguo. Para actualizarlo, seleccione "Ayuda" y, a continuación, "Eliminar modelos de aprendizaje automático" en la barra de menús. - + Layout Reset Restablecer diseño - + The window layout has been reset. The default layout will be used next time. El diseño de la ventana ha sido restablecido. La próxima vez se usará el diseño predeterminado. - + Multiple Instances Múltiples instancias - + Found process ID in lock file: ID de proceso encontrado en el archivo de bloqueo: - + Failed to delete models. Error al borrar modelos. - + Delete Failed Borrado fallido - + Models Deleted Modelos eliminados - + The models were deleted. Would you like to download them again? Los modelos se han borrado. Quieres volver a descargarlos? - + Not Supported No se admite - + Unsupported system or package format Formato de sistema o paquete no compatible - + Shutdown Apagar - + After processing, the system will shut down. Después del procesamiento, el sistema se apagará. - + You have review options enabled, these will need to be manually closed before the action can start. Tiene opciones de revisión activadas, Tiene que cerrarlas manualmente antes de comenzar. - + You have disabled writing output. Ha desactivado la escritura de salida. - + Shutting down system... Cerrando el sistema. - + Running post action "{action}"... Ejecutar la acción posterior "{action}"... - + Restart Required Reinicio necesario - + The language has been changed. Please restart the application for the changes to take effect. Se ha cambiado el idioma. Reinicie la aplicación para que los cambios surtan efecto. - - + + System Language Idioma del sistema - + Import Profile Importar perfil - + Profile Files (*.conf) Archivos de perfil (*.conf) - + Profile Imported Perfil Importado - + Import Error Error de importación - + Failed to Delete Error al borrar - + The default profile cannot be deleted. El perfil por defecto no se puede borrar. - + Delete Profile Borrar perfil - + Are you sure you want to delete the profile {profile_name}? ¿Estás seguro de que quieres borrar el perfil {profile_name}? - + Delete Error Borrar error - + Failed to delete the profile. No se ha podido eliminar el perfil. @@ -1518,217 +1518,217 @@ Switching profiles will discard changes to the current profile. Al cambiar de perfil se descartarán los cambios del perfil actual. - + Unsaved changes Cambios no guardados - - + + Load Error Error de carga - - + + Failed to load profile: Error al cargar el perfil: - + Tesseract OCR is not installed or not found Tesseract OCR no está instalado o no se encuentra. - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>No se puede'utilizar Tesseract para realizar el OCR. Volviendo a manga-ocr. Consulte las instrucciones para instalar Tesseract correctamente <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">aquí</a> o siga utilizando el modelo por defecto.</html> - + Unsupported Language Idioma no compatible. - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. El idioma '{lang_name}' no es compatible con ninguno de sus motores de OCR actuales. Consulte la documentación en línea para saber cómo añadir compatibilidad con más idiomas. - + Profile Created Perfil creado - + Create Error Crear error - + Failed to save profile to {profile_path} No se ha podido guardar el perfil en {profile_path} - - + + Save Error Guardar error - + Failed to save profile. No se ha podido guardar el perfil. - + Profile saved to {profile_path} Perfil guardado en {profile_path} - + Failed to save config. No se ha podido guardar la configuración. - + Failed to save the new profile to the configuration file. No se ha podido guardar el nuevo perfil en el archivo de configuración. - - + + No Files Sin archivos - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No hay archivos que procesar. Puede añadir archivos arrastrándolos y soltándolos en el centro de la ventana, o a través de la barra de menús: Archivos -> Añadir Archivos o Añadir Carpeta. - + No Outputs Sin salidas - + No outputs were requested. Please select at least one output before cleaning. No se ha solicitado ninguna salida. Por favor, seleccione al menos una salida antes de la limpieza. - - + + cleaned limpiado - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No hay archivos para procesar. Para editar un archivo de salida OCR antiguo, primero deberá cargar (una o más de) las imágenes a las que corresponde. - + Open OCR Output File Abrir archivo de salida OCR - + OCR Output Files (*.txt *.csv) Archivos de salida OCR (*.txt *.csv) - + Parse Error Error de análisis - + Failed to parse the OCR output file. Error al analizar el archivo de salida OCR. - + File Exists El fichero existe - + The file '{output_path}' already exists. Overwrite? El archivo '{output_path}' ya existe. ¿Sobrescribir? - + Export Images Exportar imágenes - + Would you like to export the cleaned images? ¿Desea exportar las imágenes limpiadas? - + Export OCR Results Exportar resultados de OCR - + After processing, the "{action}" action will be executed. Después de procesar, se ejecutará la acción "{action}". - - + + Post action canceled. Acción posterior cancelada. - + Un-Applied Changes Cambios no aplicados - + You have un-applied changes to the profile. Continue anyway? Tienes cambios no aplicados al perfil. ¿Continuar de todos modos? - + Would you like to export the OCR results? ¿Desea exportar los resultados del OCR? - + Processing Finished Procesamiento finalizado - + Finished processing all files. Terminado el procesamiento de todos los archivos. - + Processing Aborted Procesamiento abortado - + Processing aborted. Procesamiento abortado. - + Processing Error Error de procesamiento - + Encountered an error while processing files. Se ha producido un error al procesar los archivos. @@ -1739,7 +1739,7 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con - + Profile Perfil @@ -1754,360 +1754,365 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con Aplicar - - + + Save Guardar - + Reset All Restablecer todo - + Drag and Drop Images or Folders Here Arrastrar y soltar imágenes o carpetas aquí - + Images Imágenes - - + + File Archivo - + Size Tamaño - + Processing Size The size that the picture is shrunk down to for processing. Tamaño de procesamiento - + File Size Tamaño del archivo - + Color Mode Like RGB, grayscale etc. Modo de color - + Analytics Can also call this statistics. Analítica - + Process Proceso - + Cleaning Limpieza - + OCR OCR - + Edit Existing Output Editar salida existente - + Abort Abortar - + Start Inicio - + Output Salida - + Cleaned Image Imagen limpia - + Mask Máscara - + Isolated Text Texto aislado - - + + Review Output Revisar salida - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>Si se selecciona esta opción, se abrirá una ventana de revisión al finalizar el proceso. Los resultados también se pueden ver previamente en la vista de detalles de cada imagen', antes de proceder a guardarlos en el disco.</p></body></html> - + Save Output Guardar salida - + Output Directory: Directorio de salida: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. Puede utilizar una ruta relativa para crear una subcarpeta en la ubicación original de la imagen', o utilizar una ruta absoluta. - + Plain Text Texto sin formato - + When checked, a review window will open upon process completion. Si está marcada, se abrirá una ventana de revisión al finalizar el proceso. - + Output File: Fichero de salida: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Introduzca una ruta absoluta o relativa con un nombre de archivo. La salida de todas las imágenes se escribe en el mismo archivo.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. texto_detectado.txt - + Cancel Action Cancelar acción - + Close Cerrar - + Generating: This is present progressive, as in "[currently] Generating:" Generando: - + Current Step: Paso actual: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Imágenes procesadas: - + Set Default Fijar por defecto - + Settings Ajustes - + Theme As in color theme Tema - + Language Idioma - + Help Ayuda - + Add Files... Añadir archivos... - + Remove All Files Eliminar todos los archivos - + Remove File Eliminar archivo - + Help Translate Panel Cleaner Ayuda Traducir Panel Cleaner - + Report an Issue... Informar de un problema... - + Simulate Exception Simular excepción - + Add File Browser Extension... Añadir extensión de navegador de archivos... - + Show Supported OCR Languages Mostrar Idiomas OCR compatibles - - + + Reset Window Layout Preferences Restablecer preferencias de diseño de ventana - + Show Out Of Memory Warnings Mostrar avisos de memoria insuficiente - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Acciones post-ejecución - + Configure Post-Run Actions... Configurar acciones posteriores a la ejecución... - + New Nuevo - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>Cuando está marcada, las salidas se guardan en el disco. De lo contrario, sólo podrá previsualizarlas en la vista de detalles de la imagen. </p><p>Si el perfil permanece inalterado tras la limpieza sin esta opción activada, puede exportarlos rápidamente ejecutando de nuevo el limpiador con esta opción activada.</p></body></html> - + CSV Spreadsheet Hoja de cálculo CSV - + Delete Borrar - + Import... Importar... - + Online Documentation Documentación en línea - + About Panel Cleaner Acerca de Panel Cleaner - + View License Ver licencia - + Save as... Guardar como... - + System Sistema - + Dark Oscuro - + Light Luz - + Add Folder... Añadir carpeta... - + Delete Machine Learning Models Borrar modelos de aprendizaje automático - + Download Machine Learning Models Descargar modelos de aprendizaje automático - + Donate Donar @@ -2139,19 +2144,19 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con Modelo OCR: - + Already downloaded Ya descargado - + Text Detector model (CUDA) Modelo de detector de texto (CUDA) - + Text Detector model (CPU) Modelo de detector de texto (CPU) @@ -2170,7 +2175,7 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con Errores de descarga de OCR - + ETA estimated time of completion ETA @@ -2748,7 +2753,7 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con El nombre "{shutdown_command}" está reservado. - + The Flatpak sandbox will likely prevent actions from working. Es probable que el sandbox de Flatpak impida que las acciones funcionen. @@ -2829,73 +2834,73 @@ Consulte la documentación en línea para saber cómo añadir compatibilidad con Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input La imagen original con la escala aplicada (si es necesario). - + Input Step name in the image details view Entrada - + Input Output name in the image details view Entrada - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask La máscara aproximada generada por la IA. - + Text Detection Step name in the image details view Detección de texto - + AI Mask Output name in the image details view Máscara AI - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes Los datos de la caja sin filtrar generados por la IA. - + Raw Boxes Output name in the image details view Cajas en bruto - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes Los contornos de los cuadros de texto que ha encontrado la IA. - + Preprocessor Step name in the image details view Preprocesador - + Initial Boxes Output name in the image details view Cajas iniciales - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2903,181 +2908,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (finales). Azul: recuadros de referencia para la eliminación de ruido. - + Final Boxes Output name in the image details view Cajas finales - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask La máscara de las cajas fusionadas. - + Masker Step name in the image details view Enmascarador - + Box Mask Output name in the image details view Caja Máscara - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask La máscara de detección de texto en bruto con todo fuera de la máscara de caja recortada. - + Cut Mask Output name in the image details view Máscara de corte - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers Los diferentes pasos de crecimiento alrededor de la máscara de corte se muestran en diferentes colores. - + Mask Layers Output name in the image details view Capas de máscara - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality La desviación estándar (σ) y el grosor del contorno (en píxeles) de cada una de las mejores máscaras elegidas, si las hay. Cuanto menor sea σ, mejor, de perfecto (morado) a fallido (rojo). - + Fitment Quality Output name in the image details view Calidad de ajuste - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay La imagen de entrada con la máscara final superpuesta en color. - + Mask Overlay Output name in the image details view Máscara de superposición - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask La colección de máscaras para cada burbuja que mejor se adapte. - + Final Mask Output name in the image details view Máscara final - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text La capa de texto aislada de la imagen de entrada. - + Isolated Text Output name in the image details view Texto aislado - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output La imagen de entrada con la máscara final aplicada. - + Masked Output Output name in the image details view Salida enmascarada - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask Las máscaras que requirieron eliminación de ruido, que se superpondrán a la máscara final al exportar. - + Denoiser Step name in the image details view Denoise - + Denoise Mask Output name in the image details view Máscara de ruido - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output La imagen de entrada con la máscara final aplicada. - + Denoised Output Output name in the image details view Salida digitalizada - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask Las secciones despintadas alrededor del texto que se limpiaron mal, si es que se limpiaron. - + Inpainter Step name in the image details view Pintor - + Inpainted Mask Output name in the image details view Máscara pintada - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output La imagen de entrada: limpiada, eliminación de ruido (si está activada) y pintada. - + Inpainted Output Output name in the image details view Salida pintada - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <Esto es solo un marcador de posición, la traducción de esta cadena no es necesario> - + Write Output Output name in the image details view Salida de escritura @@ -3086,1141 +3091,1141 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin Profile - + General Profile section title General - + About this profile: Profile comment in General Acerca de este perfil: - + Notes Profile entry in General Notas - + Preferred file type to save the cleaned image as. Profile comment in General Tipo de archivo preferido para guardar la imagen limpiada. - + Preferred File Type Profile entry in General Tipo de archivo preferido - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Tipo de archivo preferido para guardar la máscara. Solo se admiten formatos de imagen que permitan transparencia. - + Preferred Mask File Type Profile entry in General Tipo de archivo de máscara preferido - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General A continuación se indican los objetivos inferior y superior para la altura de la imagen de entrada. Sólo se reduce para ajustarse al rango, prefiriendo factores numéricos enteros para minimizar el impacto en la calidad de la imagen. Las imágenes más pequeñas que cualquiera de los objetivos permanecerán inalteradas. Puede desactivar esta función estableciendo uno o ambos valores menores o iguales a 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General Esto es útil para acelerar significativamente el procesamiento de imágenes de gran tamaño. Además, como otras opciones que dependen de las dimensiones en píxeles dependen del tamaño, esto ayudará a normalizar los resultados en imágenes de distintos tamaños. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General La imagen se reduce, se procesa y, a continuación, sólo se vuelve a ampliar la máscara. Esto significa que la salida limpiada seguirá utilizando la imagen original, sin escalar, para evitar cualquier pérdida de calidad. Sólo se utiliza la altura de la imagen para determinar el factor de escala, preservando la relación de aspecto, e ignorando la anchura individual de una imagen para que el factor siga siendo coherente si una de las páginas es una doble página. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General Por ejemplo, para un objetivo inferior de 1000 y un objetivo superior de 2000, una imagen con un tamaño de 5000x7000 (an, al) píxeles se reducirá en un factor de 4, de modo que tenga un tamaño de 1250x1750 píxeles durante el procesamiento. - + Input Height Lower Target Profile entry in General Altura de entrada Objetivo inferior - + Input Height Upper Target Profile entry in General Altura de entrada Objetivo superior - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Divida tiras largas en páginas individuales. Si se activa, en lugar de apretar la tira entera para caber en la altura preferida, el clip se dividirá en páginas individuales, cada ajuste de la altura preferida más/menos doble el margen de tolerancia dividida (si el segmento estaba en medio de la raya). - + Split Long Strips Profile entry in General Dividir Tiras Largas - + Preferred height to split long strips at. Profile comment in General Altura preferida para dividir tiras largas. - + Preferred Split Height Profile entry in General Altura de división preferida - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Margen de tolerancia para dividir tiras largas, es la diferencia máxima entre la altura de división preferida y la altura real de la raya. Un algoritmo determina el mejor punto de división dentro de este margen, en un esfuerzo por evitar dividirse en medio de un panel. - + Split Tolerance Margin Profile entry in General Margen de tolerancia de división - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Relación de aspecto a utilizar para dividir tiras largas. Es la relación entre la anchura y la altura de la imagen. Si la relación de aspecto de la imagen'es menor que este valor, se considerará una tira larga.<br> Ejemplo: 0,3 significa que la anchura de la imagen es 0,3 veces la altura. - + Long Strip Aspect Ratio Profile entry in General Relación de aspecto de Tira larga - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Combinar tiras largas en una sola imagen. Si se activa, las páginas individuales creadas a partir de un clip largo se fusionarán de nuevo en una sola imagen al exportar. - + Merge After Split Profile entry in General Fusionar después de dividir - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Número máximo de hilos a usar para exportar imágenes. Puede dejarlo sin especificar para utilizar todos los hilos disponibles. Reduzca este valor si tiene problemas de memoria, que aparecerán como bloqueos aleatorios. - + Max Threads Export Profile entry in General Exportación máxima de hilos - + Text Detector Profile section title Detector de texto - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Ruta al modelo de detección de texto, dejar vacío para utilizar el modelo incorporado. Sólo se recomienda anular esto si la versión descargada automáticamente es anterior a la última versión. La ruta debe apuntar directamente al archivo comictextdetector.pt (CUDA) o comictextdetector.pt.onnx (CPU). Puedes descargar versiones anteriores del modelo <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">aquí.</a> - + Model Path Profile entry in Text Detector Modelo de ruta - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Número de modelos a ejecutar en paralelo. Esto es útil si tienes suficiente RAM (o VRAM con CUDA) para ejecutar varios modelos al mismo tiempo. Esto, por supuesto, aumentará la velocidad del proceso, pero también puede bloquear su ordenador si sobreestima su hardware. Recomiendo usar 1 modelo por cada 2 GB de memoria disponible, o 1 modelo por cada 50 imágenes. Nota: Esto se ignora si se procesan menos de 50 archivos debido a que la sobrecarga de iniciar múltiples modelos no merece la pena por debajo de esa cantidad. Advertencia: Esto puede causar inestabilidad en el programa, úselo bajo su propio riesgo. <br>¡NO informe sobre problemas con esta configuración, ya que'depende completamente del hardware! - + Concurrent Models Profile entry in Text Detector Modelos concurrentes - + Preprocessor Profile section title Preprocesador - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Los tamaños de las cajas se expresan en el número total de píxeles, por lo que una caja de 200x200 píxeles tiene un tamaño de 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Tamaño mínimo de cualquier caja para guardarla. - + Box Min Size Profile entry in Preprocessor Caja Tamaño mín. - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Tamaño mínimo de un recuadro con "lenguaje desconocido" para mantenerlo. Este lenguaje se suele asignar a logotipos y otros elementos de texto que se comportan mal. - + Suspicious Box Min Size Profile entry in Preprocessor Caja sospechosa Tamaño mín - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor El solapamiento mínimo entre dos cajas para fusionarlas. Este porcentaje (0-100) indica qué parte de la caja más pequeña debe estar dentro de la caja más grande para fusionarse. Un valor más alto requerirá un solapamiento mayor para fusionar las cajas. - + Box Overlap Threshold Profile entry in Preprocessor Umbral de solapamiento de cajas - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Si se utiliza OCR para detectar las casillas que no merece la pena limpiar', como las que sólo contienen números o símbolos. - + OCR Enabled Profile entry in Preprocessor OCR activado - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Si se utiliza Tesseract para realizar tareas de OCR.<br> Si está marcada, se puede utilizar Tesseract OCR para la extracción de texto, si está disponible.<br> Si no está marcada, se utiliza siempre el modelo OCR incorporado (manga-ocr), que es el más adecuado para el texto japonés vertical. - + OCR Use Tesseract Profile entry in Preprocessor OCR Utilizar Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor El idioma a utilizar para las tareas de OCR. El detector de texto puede detectar japonés e inglés, para cualquier otra cosa, seleccione un idioma explícitamente aquí. La detección por cuadro retiene lo que el detector de texto detectó, mientras que por página asignará el idioma más prominente a toda la página.<br> Nota: Sólo el japonés es compatible de fábrica, todo lo demás requiere que Tesseract esté habilitado y que se instalen sus paquetes de idiomas asociados. - + OCR Language Profile entry in Preprocessor Idioma OCR - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Especifica qué motor utilizar para realizar el OCR.<br> - auto: Selecciona automáticamente el motor de OCR en función del idioma detectado de cada bloque de texto dentro de la imagen. Utiliza Manga Ocr para texto japonés, Tesseract para inglés o texto desconocido.<br> - mangaocr: Obliga a Panel Cleaner a utilizar el modelo manga-ocr incorporado para todas las tareas de reconocimiento de texto. Más adecuado para texto japonés vertical.<br> - tesseract: Obliga a Panel Cleaner a utilizar Tesseract OCR para todas las tareas de reconocimiento de texto. Se trata de una opción versátil que admite el inglés y muchos otros idiomas. - + OCR Engine Profile entry in Preprocessor Motor OCR - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Define el orden de lectura para procesar y ordenar los cuadros de texto en toda la página, no en bloques de texto individuales. Esta configuración global influye en cómo se ordenan y presentan los cuadros de texto para su posterior procesamiento.<br> - auto: Detecta el orden de lectura basándose en el idioma detectado de cada bloque de texto dentro de la página.<br> - manga: Orden de derecha a izquierda, de arriba a abajo. Adecuado para manga japonés.<br> - cómic: De izquierda a derecha, de arriba abajo. Adecuado para cómics y textos occidentales.<br> Elija en función del diseño predominante de su contenido. - + Reading Order Profile entry in Preprocessor Orden de lectura - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Tamaño máximo de una caja sobre la que realizar el OCR. Estas cajas inútiles suelen ser pequeñas, y el OCR es lento, así que úsalo como límite. - + OCR Max Size Profile entry in Preprocessor OCR Tamaño máximo - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Patrón Regex para comparar con los resultados del OCR. Todo lo que coincida con este patrón se descarta. Tenga en cuenta que el modelo OCR devuelve caracteres de ancho completo, por lo que este patrón debería coincidir con ellos. - + OCR Blacklist Pattern Profile entry in Preprocessor Patrón de lista negra OCR - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor El modelo OCR sólo puede manejar texto en japonés, por lo que cuando se activa la opción estricta, descartará las casillas que no'confíe que son japonesas. A veces, los números u otros símbolos disminuyen su confianza, lo que hace que el idioma detectado sea desconocido. Si se desactiva la opción estricta, no se descartarán. Todo lo que se reconozca con seguridad como un idioma diferente se descartará independientemente de esta configuración. - + OCR Strict Language Profile entry in Preprocessor OCR Lenguaje estricto - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Relleno para añadir a cada lado de una caja. Esto se añade a las cajas iniciales creadas por el detector de texto AI. Estas cajas se visualizan en verde. - + Box Padding Initial Profile entry in Preprocessor Caja Relleno Inicial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Relleno para añadir a la derecha de una caja. Esta extensión ayuda a cubrir el texto ruby que flota a la derecha del texto vertical. - + Box Right Padding Initial Profile entry in Preprocessor Caja derecha Relleno inicial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Relleno para añadir a cada lado de una caja. Esto se añade a un conjunto ampliado de recuadros, utilizados para eliminar los falsos positivos por el detector de texto AI's mask. Estos recuadros se visualizan en púrpura. - + Box Padding Extended Profile entry in Preprocessor Acolchado de cajas ampliado - + Box Right Padding Extended Profile entry in Preprocessor Caja derecha ampliada - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Relleno para añadir a cada lado de una caja. Esto se añade a las cajas de referencia utilizadas para muestrear la imagen original mientras se analiza qué máscara utilizar. Estas cajas se visualizan en azul. - + Box Reference Padding Profile entry in Preprocessor Relleno de referencia de caja - + Masker Profile section title Enmascarador - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Número máximo de hilos a usar para la generación de máscaras. Puede dejarlo sin especificar para utilizar todos los hilos disponibles. Bajar este valor si se encuentran problemas de memoria, que aparecerán como bloqueos aleatorios. - + Max Threads Profile entry in Masker Hilos máximos - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Número de píxeles para hacer crecer la máscara en cada paso. Esto abulta el contorno de la máscara, por lo que valores más pequeños serán más precisos pero más lentos. - + Mask Growth Step Pixels Profile entry in Masker Máscara Crecimiento Paso Píxeles - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Número de pasos para hacer crecer la máscara. Un número mayor hará más máscaras y más grandes, limitadas en última instancia por el tamaño de la caja de referencia. - + Mask Growth Steps Profile entry in Masker Pasos de crecimiento de la máscara - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Espesor mínimo de una máscara. Esto es como el primer paso de crecimiento de la máscara', el resto seguirá el Paso de crecimiento de la máscara Píxeles. De esta manera, usted puede tener un pequeño tamaño de paso para la precisión, pero aún así evitar máscaras muy finas que podrían ser utilizados para limpiar el texto sólo rodeado por un contorno, cuando inpainting've sido la mejor opción. - + Min Mask Thickness Profile entry in Masker Espesor mínimo de la máscara - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter La desviación estándar mínima de colores alrededor del borde de una máscara determinada para realizar pintura en la región alrededor de la máscara. Si este valor coincide con la desviación máxima para las máscaras, sólo se pintarán burbujas fallidas, haciendo que los siguientes dos ajustes sean irrelevantes. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine las salidas en un solo archivo de proyecto como capas. Formatos soportados actualmente: Photoshop PSD.<br> - ninguno: Cada imagen y máscara se guardan como archivos básicos.<br> - psd-per-imagen: Las imágenes y máscaras se guardan juntas en un archivo PSD por imagen de entrada.<br> - psd-bulk: Todas las imágenes y máscaras se guardan juntas en un único archivo PSD, agrupado por imagen de entrada. - + Layered Export Profile entry in General Exportación por capas - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Si se permiten máscaras de color. Cuando sea verdadero, el enmascarador permitirá que las máscaras usen cualquier color, no solo blanco, negro o gris. - + Allow Colored Masks Profile entry in Masker Permitir máscaras de colores - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Umbral máximo para que un píxel se considere blanquecino. El color medio a lo largo del borde de una máscara puede no ser blanco puro, por lo que para evitar ligeros tonos blanquecinos sobre un fondo blanco puro, cualquier valor más claro que este umbral se redondeará a blanco puro. Rango de valores: negro (0) a blanco puro (255). - + Off White Max Threshold Profile entry in Masker Off White Umbral máximo - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker La desviación estándar de una máscara representa la variación de color a lo largo del borde de la máscara. Para ello, sólo se muestrea de la imagen original la única línea de píxeles a lo largo del borde exterior de una máscara. Una desviación baja significa que la máscara se sitúa en un color sólido, lo que significa que'no se cruza con ningún texto u otros objetos. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Mejora mínima de la desviación estándar de la máscara para seguir haciéndola crecer. Establecer un valor más alto aquí requiere una mejora mayor para considerar una máscara más grande, para dar preferencia a las máscaras más pequeñas. - + Mask Improvement Threshold Profile entry in Masker Umbral de mejora de la máscara - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Si se utiliza el algoritmo de selección rápida de máscaras. Cuando es verdadero, el algoritmo de selección de máscara elegirá la primera máscara perfecta, si se encuentra una pronto. Esto es más rápido, pero puede no encontrar la mejor máscara, si una ligeramente mayor hubiera sido mejor. - + Mask Selection Fast Profile entry in Masker Selección rápida de máscaras - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Número máximo de hilos a utilizar para la eliminación de ruido. Puede dejarlo sin especificar para utilizar todos los subprocesos disponibles. Reduzca este valor si tiene problemas de memoria, que aparecerán como bloqueos aleatorios. - + Max Threads Profile entry in Denoiser Máximo de hilos - + Inpainter Profile section title Pintor - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter CARACTERÍSTICA EXPERIMENTAL: Si encuentra una configuración predeterminada mejor, abra una incidencia en github para compartir las mejoras con todos. Tenga en cuenta que inpainting no está'habilitado por defecto. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting es cuando se utiliza el aprendizaje automático para sustituir el contenido de una imagen basándose en su entorno.'Para las máscaras que no se han podido limpiar bien (o no se han podido limpiar en absoluto), se puede utilizar el inpainting. Para diferenciarlo del denoising, el inpainting está pensado para máscaras significativamente peores que se ajustaron bien. Las máscaras a las que se les haya aplicado la eliminación de ruido no se pintarán.' - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Como este paso puede dar malos resultados en algunos casos, puede desactivarse aquí. Desmarque esta opción para desactivar el repintado. - + Inpainting Enabled Profile entry in Inpainter Inpainting Activado - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter El radio máximo de una máscara sobre el que realizar el inpainting. Las máscaras mayores que esto se dejarán como están, porque si el margen es tan grande, es probable que la máscara ya sea suficientemente buena. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter Radio mínimo alrededor de la máscara que se va a pintar. Se añade al tamaño óptimo de la máscara para garantizar que el repintado cubra toda la máscara. - + Min Inpainting Radius Profile entry in Inpainter Radio mínimo de pintado - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter En el caso de las máscaras que resultan mucho más difíciles de limpiar, es decir, que tienen una desviación estándar elevada, se aumenta el radio del inpainting para cubrir más superficie de la máscara. Este margen adicional se añade al radio mínimo de repintado y se calcula de la siguiente manera: multiplicador del radio de repintado por la desviación típica de la máscara. - + Inpainting Radius Multiplier Profile entry in Inpainter Multiplicador del radio de pintado - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter El radio máximo alrededor de la máscara a pintar. Esto limita el tamaño al que puede crecer el pintado, para evitar que cubra demasiada imagen si se utiliza un multiplicador de radio grande. - + Max Inpainting Radius Profile entry in Inpainter Radio máximo de pintado - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter Tras el repintado, recorte el resultado de la imagen original para evitar que el repintado afecte al resto de la imagen. De este modo, se preserva al máximo la imagen original. Este radio se añade alrededor del radio final de inpainting, debido a que el modelo de inpainting modifica algunos píxeles fuera de su región dedicada. - + Inpainting Isolation Radius Profile entry in Inpainter Radio de aislamiento - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Desvanezca los bordes de la máscara de cobertura pintada en esta cantidad de píxeles para mezclar suavemente las partes pintadas con el resto de la imagen.<br> Si ve contornos tenues después de pintar, aumente el valor de pintado mínimo y establezca éste a la mitad de esa cantidad. - + Inpainting Fade Radius Profile entry in Inpainter Radio de desvanecimiento - + Detect per box Language option for OCR Detectar por caja - + Detect per page Language option for OCR Detectar por página - + Japanese Language option for OCR Japonés - + English Language option for OCR Inglés - + Korean Language option for OCR Coreano - + Korean (vertical) Language option for OCR Coreano (vertical) - + Chinese - Simplified Language option for OCR Chino simplificado - + Chinese - Traditional Language option for OCR Chino tradicional - + Albanian Language option for OCR Albano - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengalí - + Bulgarian Language option for OCR Búlgaro - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalán; Valenciano - + Croatian Language option for OCR Croata - + Czech Language option for OCR Checo - + Danish Language option for OCR Danés - + Dutch; Flemish Language option for OCR Neerlandés; Flamenco - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finlandés - + French Language option for OCR Francés - + Georgian Language option for OCR Georgiano - + German Language option for OCR Alemán - + Greek Language option for OCR Griego - + Hebrew Language option for OCR Hebreo - + Hindi Language option for OCR Hindú - + Hungarian Language option for OCR Húngaro - + Indonesian Language option for OCR Indonesio - + Italian Language option for OCR Italiano - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latín - + Lithuanian Language option for OCR Lituano - + Luxembourgish Language option for OCR Luxemburgués - + Malay Language option for OCR Malayo - + Mongolian Language option for OCR Mongol - + Nepali Language option for OCR Nepalí - + Norwegian Language option for OCR Noruego - + Persian Language option for OCR Persa - + Polish Language option for OCR Polaco - + Portuguese Language option for OCR Portugués - + Romanian; Moldavian Language option for OCR Rumano; Moldavo - + Russian Language option for OCR Ruso - + Serbian Language option for OCR Serbio - + Serbian - Latin Language option for OCR Serbio - Latín - + Slovak Language option for OCR Eslovaco - + Slovenian Language option for OCR Eslovaco - + Spanish; Castilian Language option for OCR Español, Castellano - + Swedish Language option for OCR Sueco - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Tailandés - + Turkish Language option for OCR Turco - + Ukrainian Language option for OCR Ucraniano - + Vietnamese Language option for OCR Vietnamita - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker La desviación estándar máxima de una máscara a considerar. Un valor alto aquí significa una mayor tolerancia para la máscara de intersección de texto u otros objetos, que no es'una buena máscara, ya que requerirá inpainting de todos modos. Un valor de 0 sólo permite máscaras perfectas, lo que se recomienda para imágenes de muy alta resolución. - + Mask Max Standard Deviation Profile entry in Masker Máscara Desviación estándar máxima - + Color to use for the debug mask. Profile comment in Masker Color a utilizar para la máscara de depuración. - + Debug Mask Color Profile entry in Masker Color de la máscara de depuración - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser Cuando un poco de texto es difícil de enmascarar, el paso de limpieza probablemente tuvo que elegir una máscara pequeña, que deja un montón de artefactos jpeg detrás, si estaban alrededor del texto. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser Esto significa que el color de los píxeles alrededor del borde de una máscara no es'uniforme, lo que se cuantifica como desviación estándar. El denoiser puede centrarse en máscaras con una desviación estándar mínima y denoise el área justo alrededor de la máscara. De este modo, se conservan los detalles del resto de la imagen, pero se eliminan los artefactos justo alrededor del texto. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Dado que se trata de un paso opcional e incluso puede ser superfluo para imágenes de alta resolución que no'sufren de jpeg-artifacts, se puede desactivar aquí. Seleccione False para desactivar la eliminación de ruido. - + Denoising Enabled Profile entry in Denoiser Eliminación de ruido activada - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser La desviación estándar mínima de los colores alrededor del borde de una máscara dada para realizar la eliminación de ruido en la región alrededor de la máscara. - + Noise Min Standard Deviation Profile entry in Denoiser Ruido Desviación estándar mín - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Nota: Si se activa el inpainting, la desviación estándar mínima del inpainting actuará como máximo para esto, después de lo cual esta máscara se entrega al inpainter - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser El grosor de un contorno para denotar alrededor de una máscara. - + Noise Outline Size Profile entry in Denoiser Ruido Tamaño del contorno - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Desvanezca los bordes de la máscara de cobertura desprotegida en esta cantidad para mezclar suavemente las partes desprotegidas con el resto de la imagen. - + Noise Fade Radius Profile entry in Denoiser Radio de atenuación del ruido - + Set to True to support colored images. Profile comment in Denoiser Establézcalo a True para admitir imágenes en color. - + Colored Images Profile entry in Denoiser Imágenes en color - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser La intensidad de la eliminación de ruido de la imagen. Los valores más altos significan más eliminación de ruido. - + Filter Strength Profile entry in Denoiser Resistencia del filtro - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser La intensidad de la eliminación de ruido de los canales de color, en contraposición a la luminosidad. Los valores más altos significan más eliminación de ruido. - + Color Filter Strength Profile entry in Denoiser Intensidad del filtro de color - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Tamaño en píxeles del parche de plantilla que se utiliza para calcular los pesos. Debe ser impar. - + Template Window Size Profile entry in Denoiser Tamaño de la ventana de plantilla - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Tamaño en píxeles de la ventana que se utiliza para calcular la media ponderada de un píxel determinado. Debe ser impar. - + Search Window Size Profile entry in Denoiser Tamaño de la ventana de búsqueda @@ -4247,7 +4252,7 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. Todos los Núcleos @@ -4664,14 +4669,14 @@ Verde: cajas iniciales. Rojo: cajas ampliadas. Morado: recuadros fusionados (fin ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>La ruta del modelo del Detector de Texto {value} no existe, volviendo al modelo por defecto. Puede descargar el modelo manualmente desde <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">aquí</a> o seguir utilizando el modelo por defecto.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. El patrón regex "{value}" para {key} no es válido, volviendo al predeterminado. From 643193d9fbf10aa09c05de9077b23c8a7a5daae2 Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:10 +0100 Subject: [PATCH 07/18] New translations panelcleaner_source.ts (Bulgarian) --- translations/PanelCleaner_bg_BG.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_bg_BG.ts b/translations/PanelCleaner_bg_BG.ts index c591d7f..18680a0 100644 --- a/translations/PanelCleaner_bg_BG.ts +++ b/translations/PanelCleaner_bg_BG.ts @@ -4,7 +4,7 @@ - + default reserved profile name по подразбиране @@ -20,54 +20,54 @@ Профил {profile_name} създаден. - + Profile name cannot be empty. Името на профила не може да е празно. - + Profile name already in use. Името на профила вече се използва. - + Profile name is reserved. Профилното име е резервирано. - - - - + + + + Unknown Неизвестно - + RGB Color mode РДЖБ - + CMYK Color mode ЦМЖЧ - + Grayscale Color mode Скала на нивата на сиво - + 1-bit Color mode 1-бит - + Palette Color mode Палитра @@ -78,45 +78,45 @@ Грешка при изтегляне на файла от адрес: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Грешки при изтегляне на файла от адрес: {url} Неуспех при запис на файла към {save_path} - + Error downloading file from url: {url} The file content is different from expected. Грешка при изтеглянето на файл от Url: {url} Съдържанието на файла е различно от очакваното. + - - - + + + - - - + + N/A Няма/Не е приложимо - + OCR Analytics ОРЗ Анализ - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Брой кутии: {num_boxes} | Брой на малки кутии: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Брой на отстранени кутии: {num_removed} ({removed_box_ratio} общоl, {removed_among_small_ratio} малки кутии) @@ -155,66 +155,66 @@ Removed bubbles: - + Small boxes Малки кутии - + Removed boxes Отстранени кутии - - + + Failed Провал - + Perfect Перфектно - + Total Общо - + Mask Fitment Analytics Анализ на пригодността на маската - + Total Общо - + Total boxes Общ брой кутии - + Masks succeeded Успешни маски - + Masks failed Провалени маски - + Perfect masks Отлични маски - + Average border deviation Средно гранично отклонение @@ -233,7 +233,7 @@ Mask usage by thickness (in pixels): Маска ({thickness}px) - + Box mask Маска на кутията @@ -254,27 +254,27 @@ Pages with failures / total: - + Denoising Analytics Анализ на отстраняване на шум - + Total masks Сума на маските - + Masks denoised Маски с понижено ниво на шум - + Minimum deviation to denoise Минимално отклонение за денонозация - + Maximum allowed deviation Максимално допустимо отклонение @@ -293,27 +293,27 @@ Pages with failures / total: Шумо‑премахнат - + Inpainting Analytics Анализ на обрисуване - + Inpainting performed Извършено обрисуване - + Average thickness Средна дебелина - + Minimum thickness Минимална дебелина - + Maximum thickness Максимална дебелина @@ -372,7 +372,7 @@ Pages with failures / total: The file format was not recognized. - + The CSV file must start with a header row, followed by data rows. The CSV file must start with a header row, followed by data rows. @@ -402,7 +402,7 @@ Pages with failures / total: Multiple images match the file path: {path} - + Error, no version found. Error, no version found. @@ -412,7 +412,7 @@ Pages with failures / total: CUDA Error - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pages with failures / total: Модел по обрисуване не бе открит. - + filename име на файла - + startx startx - + starty Starty - + endx endx - + endy endy - + text текст - + Saved detected text to {output_file} Запазване на открития текст в {output_file} - + Failed to write detected text to {output_file} Не успя да запише открития текст в {output_file} - + Save Failed Неуспех при Запис - + Failed to write detected text to file. Не успя да запише открития текст във файла. @@ -538,7 +538,7 @@ Pages with failures / total: - + Page layered export @@ -624,7 +624,7 @@ An AI-powered tool to clean manga panels. ColorButton - + Select Color Избери цвят @@ -670,7 +670,7 @@ An AI-powered tool to clean manga panels. Неподдържана система - + This system is not supported. Тази система не се поддържа. @@ -685,12 +685,12 @@ An AI-powered tool to clean manga panels. Деинсталирай за Делфин - + The extension will be installed at {path}. This doesn't require root permissions. Добавката ще бъде инсталирана в {path}. Това не изисква права на админ. - + The extension file will be deleted. This doesn't require root permissions. Файла‑добавка ще бъде премахнат. Това не изисква права на админ. @@ -705,17 +705,17 @@ An AI-powered tool to clean manga panels. Деинсталиране за Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. Добавката ще бъде инсталирана в регистъра на Windows.Това изисква администраторски права. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. Добавката ще бъде деинсталирана в регистъра на Windows.Това изисква администраторски права. - + Dolphin extension installed successfully. Делфин разширението беше успешно инсталирано. @@ -725,7 +725,7 @@ An AI-powered tool to clean manga panels. Инсталирането успешно - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Делфин разширението се инсталира успешно. @@ -805,12 +805,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) Оригинален размер в пиксели (ширина × височина) - + Processing size in pixels (width × height), scale factor Размер на обработката в пиксели (ширина × височина), фактор на мащаба @@ -836,7 +836,7 @@ By default, this uses the built-in profile. However, you can change this default Неподдържани ТИФФ файлове - + The following 5-channel TIFF files are not supported: Следните 5-канални ТИФФ файла не се поддържат: @@ -863,27 +863,27 @@ By default, this uses the built-in profile. However, you can change this default split - + Number of boxes removed by the OCR model / total boxes Брой на кутиите премахнати чрез модел ОРЗ / сума на кутии - + Number of boxes that failed to generate a mask / total boxes Брой на кутиите които не успяха да генерират маска / сума на кутии - + Number of boxes that were perfectly masked / total boxes Брой кутии, които са били идеално маскирани, / общ брой кутии - + Number of boxes that were denoised / total boxes Брой кутии, които бяха шумо-потиснати / сума кутии - + Number of boxes that were inpainted Брой обрисувани кутии @@ -1031,14 +1031,14 @@ By default, this uses the built-in profile. However, you can change this default Невъзможна Заявка - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Отстраняването на шума е забранено за текущия профил, този изход не може да бъде генериран. Моля разрешете отстраняване на шума в профилните настройки и опитайте отново. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Обрисуванeто е забранено за текущия профил, този изход не може да бъде генериран. @@ -1060,7 +1060,7 @@ Please enable inpainting in the profile settings and try again. Експортът е неуспешен - + Failed to export image: Провал при експортиране на изображението: @@ -1103,7 +1103,7 @@ Please enable inpainting in the profile settings and try again. result - + results results @@ -1118,17 +1118,17 @@ Please enable inpainting in the profile settings and try again. new image - + new images new images - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. @@ -1227,12 +1227,12 @@ Please enable inpainting in the profile settings and try again. Бележка: Име "{name}" бе скрит - + Corrupted log session Повредена сесия на дневника - + Current session Текуща сесия @@ -1247,7 +1247,7 @@ Please enable inpainting in the profile settings and try again. Грешка - + Errors Грешки @@ -1257,7 +1257,7 @@ Please enable inpainting in the profile settings and try again. Критично - + Criticals Критични @@ -1273,49 +1273,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button Отвори Конфигурация - + Open Log Statusbar button Отвори Лога - + CUDA Enabled КУДА Разрешен - + Select Output Directory Избери Изходна Директория - + Aborting... Преустановяване… - + Loading OCR model... Зареждане на ОРЗ модел… - + Loaded OCR model. Зареден ОРЗ модел. - + Error Грешка - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1324,7 +1324,7 @@ Continue anyway? Continue anyway? - + Delete Models Изтрий Модели @@ -1334,23 +1334,23 @@ Continue anyway? Сигурни ли сте, че искате да изтриете моделите за машинно обучение? Това ще направи невъзможно почистването и ОРЗ, докато не бъдат изтеглени отново. - + Encountered error: Възникнала грешка: - + MPS Enabled Активиран MPS - - + + Inpainting Model Update Inpainting Model Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1364,148 +1364,148 @@ Switch to the new model? Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. - + Layout Reset Layout Reset - + The window layout has been reset. The default layout will be used next time. The window layout has been reset. The default layout will be used next time. - + Multiple Instances Множество инстанции - + Found process ID in lock file: Открит процес ид в заключен файл: - + Failed to delete models. Неуспех при изтриването на моделите. - + Delete Failed Неуспешно изтриване - + Models Deleted Моделите Изтрити - + The models were deleted. Would you like to download them again? Моделите бяха изтрити. Желаете ли да ги изтеглите отново? - + Not Supported Не се поддържа - + Unsupported system or package format Неподдържана система или пакетен формат - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Необходим е рестарт - + The language has been changed. Please restart the application for the changes to take effect. Езика бе променен. Моля рестартирайте апликацията, за да влязат промените в сила. - - + + System Language Системен Език - + Import Profile Импортирай Профил - + Profile Files (*.conf) Профилни Файлове (*.conf) - + Profile Imported Профил Импортиран - + Import Error Грешка при Импортирането - + Failed to Delete Неуспех при Изтриване - + The default profile cannot be deleted. Профила по подразбиране не може да бъде премахнат. - + Delete Profile Изтрий Профил - + Are you sure you want to delete the profile {profile_name}? Сигурни ли сте че желаете да изтриете профил {profile_name}? - + Delete Error Изтрий Грешка - + Failed to delete the profile. Провал при изтриване на профила. @@ -1517,217 +1517,217 @@ Switching profiles will discard changes to the current profile. Превключването на профила ще доведе до отхвърляне на промените в текущия профил. - + Unsaved changes Незапазени промени - - + + Load Error Грешка при Зареждане - - + + Failed to load profile: Провал при зареждане на профил: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Tesseract не може да бъде използван с ОРЗ. Връщане към манга-орз. Моля, вижте инструкциите за правилно инсталиране на Tesseract тук <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr"></a> или продължете да използвате модела по подразбиране.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Профил Създаден - + Create Error Грешка Създаване - + Failed to save profile to {profile_path} Не успях да запазя профил в {profile_path} - - + + Save Error Грешка при Записване - + Failed to save profile. Провал при записване на профила. - + Profile saved to {profile_path} Профил записан в {profile_path} - + Failed to save config. Не успях да запиша конфигурацията. - + Failed to save the new profile to the configuration file. Неуспешно записване на новия профил в конфигурационния файл. - - + + No Files Няма Файлове - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. Няма файлове за обработка. Можете да добавяте файлове, като ги плъзгате и пускате в средата на прозореца или чрез менюто: Файлове -> Добавяне на файлове или Добавяне на папка. - + No Outputs Няма изходящи данни - + No outputs were requested. Please select at least one output before cleaning. Изходящи данни не бяха поискани. Моля изберете поне един изход преди почистване. - - + + cleaned почистено - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists Файла Съществува - + The file '{output_path}' already exists. Overwrite? Файла '{output_path}' вече съществува. Презапиши? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Обработката Завърши - + Finished processing all files. Обработката на всички файлове Приключи. - + Processing Aborted Обработването Прекратено - + Processing aborted. Обработката спряна. - + Processing Error Грешка при Обработването - + Encountered an error while processing files. Натъкване на грешки, при обрабтката на файлове. @@ -1738,7 +1738,7 @@ Check the online documentation for how to add support for more languages. - + Profile Профил @@ -1753,363 +1753,368 @@ Check the online documentation for how to add support for more languages.#Приложи - - + + Save Запис - + Reset All #Нулирай в-ки - + Drag and Drop Images or Folders Here #Плъзнете и Пуснете Изображения или Папки Тук - + Images Изображения - - + + File Файл - + Size Размер - + Processing Size The size that the picture is shrunk down to for processing. Големина на обработка - + File Size Файлов размер - + Color Mode Like RGB, grayscale etc. Цветови Режим - + Analytics Can also call this statistics. Статистика - + Process #Процес - + Cleaning #Почистване - + OCR #ОРЗ - + Edit Existing Output Edit Existing Output - + Abort #Прекрати - + Start #Начало - + Output #Изход - + Cleaned Image #Изчистено копие - + Mask #Маска - + Isolated Text #Изолиран Текст - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Запис на Изход - + Output Directory: Изходна Директория: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. Можете да използвате относителен път, за да създадете подпапка в оригиналното местоположение на изображението, или да използвате абсолютен път. - + Plain Text Обикновен текст - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Изходящ файл: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Въведете абсолютен или относителен път с име на файл. Резултатът от всички изображения се записва в един и същ файл.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. открит_текст.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" #Генериране: - + Current Step: #Настояща Стъпка: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. #Обработени Образи: - + Set Default Задаване по подразбиране - + Settings Настройки - + Theme As in color theme Тема - + Language Език - + Help Помощ - + Add Files... Добави Файлове… - + Remove All Files Премахни Всички Файлове - + Remove File Премахни Файл - + Help Translate Panel Cleaner Укажи Съдействие в Превода на Panel Cleaner - + Report an Issue... Докладвай проблем… - + Simulate Exception Симулирай Изключение - + Add File Browser Extension... Добавяне на разширение за браузъра на файлове... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New Нов - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Изтрий - + Import... Импорт… - + Online Documentation Онлайн Документация - + About Panel Cleaner За Panel Cleaner - + View License Преглед Лиценз - + Save as... Запис като… - + System Система - + Dark Тъмна - + Light Светла - + Add Folder... Добави Папка… - + Delete Machine Learning Models Изтриване на модели за машинно обучение - + Download Machine Learning Models Изтегляне на модели за машинно обучение - + Donate Дари @@ -2141,19 +2146,19 @@ Check the online documentation for how to add support for more languages.ОРЗ модел: - + Already downloaded Вече изтеглено - + Text Detector model (CUDA) Модел Засичане на текст (CUDA) - + Text Detector model (CPU) Модел Засичане на текст (ЦПУ) @@ -2172,7 +2177,7 @@ Check the online documentation for how to add support for more languages.Грешки при Теглене на ОРЗ - + ETA estimated time of completion Остават @@ -2750,7 +2755,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2831,73 +2836,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input Първоначалното изображение с приложено мащабиране (ако е необходимо). - + Input Step name in the image details view Вход - + Input Output name in the image details view Вход - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask Грубата маска, генерирана от ИИ. - + Text Detection Step name in the image details view Засичане на текст - + AI Mask Output name in the image details view ИИ Маска - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes Нефилтрирани данни от кутия, генерирани от ИИ. - + Raw Boxes Output name in the image details view Сурови Кутии - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes Контурите на текстовите полета, които ИИ е намерил. - + Preprocessor Step name in the image details view Предпроцесор - + Initial Boxes Output name in the image details view Първоначални Кутии - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2905,181 +2910,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Зелено: първоначални полета. Червено: разширени полета. Лилаво: обединени (крайни) полета. Синьо: референтни кутии за деноализацията. - + Final Boxes Output name in the image details view Окончателни Кутии - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask Маската на слетите кутии. - + Masker Step name in the image details view Маскиращ - + Box Mask Output name in the image details view Маска на кутията - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask Грубата маска за откриване на текст с изрязано всичко извън маската на полето. - + Cut Mask Output name in the image details view Маска за Изрязване - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers Различните етапи на растеж около маската за рязане са изобразени в различни цветове. - + Mask Layers Output name in the image details view Слоеве с маски - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality Стандартното отклонение (σ) и дебелината на контура (в пиксели) на всяка избрана най-добра маска, ако има такава. По-ниското σ е по-добро, от перфектно (лилаво) до неуспешно (червено). - + Fitment Quality Output name in the image details view Качество на прилягане - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay Входното изображение с окончателната маска, покрита с цвят. - + Mask Overlay Output name in the image details view Маска за наслагване - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask Колекцията от маски за всеки балон, която е най-подходяща. - + Final Mask Output name in the image details view Финална Маска - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text Текстовият слой, изолиран от входното изображение. - + Isolated Text Output name in the image details view Изолиран Текст - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output Входно изображение с приложена окончателна маска. - + Masked Output Output name in the image details view Маскиран изход - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask Маските, които се нуждаят от регенериране, които да се наслагват върху крайната маска при експортиране. - + Denoiser Step name in the image details view Неутрализатор на Шум - + Denoise Mask Output name in the image details view Маска на Неутрализатор - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output Входното изображение с приложената редуцираща маска. - + Denoised Output Output name in the image details view Неутрализатор Изход - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask Неизрисуваните участъци около текста, които са били лошо почистени, ако изобщо са били почистени. - + Inpainter Step name in the image details view Обрисуване - + Inpainted Mask Output name in the image details view Маска на Обрисуване - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output Входното изображение: почистено, шумо‑редуцирано (ако е разрешено) и обрисувано. - + Inpainted Output Output name in the image details view Обрисуване Изход - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3088,1141 +3093,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title Общи - + About this profile: Profile comment in General За този профил: - + Notes Profile entry in General Бележки - + Preferred file type to save the cleaned image as. Profile comment in General Предпочитан тип на файла, като какъвто да запазите почистеното изображение. - + Preferred File Type Profile entry in General Предпочитан Тип на файла - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Предпочитан тип на файла, като който да запазите маската. Поддържат се само формати на изображения, които позволяват прозрачност. - + Preferred Mask File Type Profile entry in General Предпочитан вид на маската - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General По-долу са посочени долната и горната цел за височината на входното изображение. Мащабът се намалява само за да се вмести в диапазона, като се предпочитат целочислени коефициенти, за да се сведе до минимум въздействието върху качеството на изображението. Изображения, по-малки от всяка от двете цели, ще останат непроменени. Можете да деактивирате тази функция, като зададете една или и двете стойности по-малки или равни на 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General Това е от полза, когато е нужно значително ускорение при обработката на големи изображения. В допълнение, след като други опции разчитат на пикселните размери, обвързани с техния размер, това ще спомогне за резултатите при различните размери на изображенията. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General Образът ще бъде смален, обработен и в крайна сметка единствено маската ще бъде увеличена отново. Значейки че почистения изход все още ще използва оригиналния, не мащабиран образ за да се предотврати загуба на качество. Само височината на изображението ще бъде използвано при определяне на коефициент на мащаба, запазвайки пропорциите, и игнорирайки индивидуалната ширина на изображението, така че коефициентът да остане непроменен, ако една от страниците е двойна. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General например – за долна цел от 1000 и горна цел от 2000 изображението с размер 5000x7000 (д, ш) пиксела ще бъде намалено с коефициент 4, така че по време на обработката да има размер 1250x1750 пиксела. - + Input Height Lower Target Profile entry in General Начална височина Долна цел - + Input Height Upper Target Profile entry in General Начална височина Горна цел - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Текстов детектор - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Път към модела за тексто-засичане, оставете празно за да използвате вграденият. Необходимо е презаписване ако автоматично-изтеглената версия е по-стара от най-новото издание. Пътят трябва директно да посочва комикс‑детектора.pt (КУДА) или комикс‑детектора.pt.onnx (ЦПУ) файла. Можете да изтеглите стари версии на модела оттук. <a href="https://github.com/zyddnys/manga-image-translator/releases/latest"></a> - + Model Path Profile entry in Text Detector Път на модела - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Брой на моделите, които се изпълняват паралелно. Това е полезно, ако разполагате с достатъчно RAM (или VRAM при CUDA), за да стартирате няколко модела едновременно. Това, разбира се, ще увеличи скоростта на процеса, но може и да срине компютъра ви, ако надцените хардуера си. Препоръчвам да използвате 1 модел на 2 GB налична памет или 1 модел на 50 изображения. Забележка: Това се пренебрегва, ако се обработват по-малко от 50 файла, тъй като под тази стойност не си заслужава да се стартират множество модели. Предупреждение: Това може да доведе до нестабилност на програмата, използвайте го на свой риск. <br>НЕ съобщавайте за проблеми, свързани с този проблем, тъй като той е много зависим от хардуера! - + Concurrent Models Profile entry in Text Detector Конкурентни модели - + Preprocessor Profile section title Препроцесор - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Размерите на полетата се задават в общия брой пиксели, така че поле с размери 200x200 пиксела има размер 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Минимален размер на кутията, за да се съхрани. - + Box Min Size Profile entry in Preprocessor Кутия Мин. Размер - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Минимален размер на кутия с "неизвестен" език за запазване. Езикът е най-вероятно зададен за логота и други зле представящи се текстови елементи. - + Suspicious Box Min Size Profile entry in Preprocessor Подозрителен мин. размер на кутия - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor Минималното припокриване между две полета за сливането им. Този процент (0-100) означава каква част от по-малката кутия трябва да е вътре в по-голямата, за да бъде обединена. По-висока стойност изисква по-голямо припокриване, за да се слеят кутиите. - + Box Overlap Threshold Profile entry in Preprocessor Праг на припокриване на полето - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Дали да използва ОРЗ за засичане на кутии/полета, които не си заслужава да се почистват, като например тези които съдържат само числа или символи. - + OCR Enabled Profile entry in Preprocessor ОРЗ Включено - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Дали да се използва Tesseract за изпълнение на OCR задачи.<br> Ако е поставена отметка, Tesseract OCR може да се използва за извличане на текст, ако е наличен.<br> Ако не е поставена отметка, винаги се използва вграденият OCR модел (manga-ocr), който е най-подходящ за вертикален японски текст. - + OCR Use Tesseract Profile entry in Preprocessor ОРЗ Използвай Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Определя кой енджин да бъде използван за ОРЗ.<br> - авто: Автоматично избира енджина за OCR въз основа на открития език на всеки текстов блок в изображението. Използва Manga Ocr за японски текст, Tesseract за английски или неизвестен текст.<br> - mangaocr: Принуждава програмата Panel Cleaner да използва вградения модел manga-ocr за всички задачи за разпознаване на текст. Най-подходящ за вертикален японски текст.<br> - tesseract: Изисква от Panel Cleaner да използва Tesseract OCR за всички задачи за разпознаване на текст. Това е универсална опция, която поддържа английски и множество други езици. - + OCR Engine Profile entry in Preprocessor Енджин ОРЗ - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Определя реда на четене за обработка и сортиране на текстови полета на цялата страница, а не на отделни текстови блокове. Тази глобална настройка влияе върху начина, по който текстовите полета се подреждат и поднасят за по-нататъшна обработка.<br> - авто: Определя реда на четене въз основа на открития език на всеки текстов блок в страницата.<br> - манга: Дясно-към-Ляво, от горе надолу. Подходящо за Японска манга.<br> - комикс: Ляво-към-Дясно, от горе надолу. Подходящ за Западни комикси и текстове.<br> Изберете в съответствие с преобладаващото оформление на съдържанието си. - + Reading Order Profile entry in Preprocessor Подредба на четенето - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Максимален размер на полето, върху което се извършва OCR. Тези безполезни полета обикновено са малки ала OCR е бавен, тъй че използвайте това като граница. - + OCR Max Size Profile entry in Preprocessor OCR Макс. размер - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor ОРЗ (OCR) Черен /Забранителен/ Списък ‑ Модел - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor Стриктен Език за ОРЗ - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Добавяне на уплътнение /padding/ към всяка страна на кутията. Това се добавя към първоначалните кутии, създадени от ИИ на детектора на текст. Тези кутии са визуализирани в зелено. - + Box Padding Initial Profile entry in Preprocessor Първоначално уплътнение за поле - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Уплътняването, което да се добави към дясната страна на кутия. Разширението помага за покриване на текст‑руби, който плава в дясно при вертикален текст. - + Box Right Padding Initial Profile entry in Preprocessor Вдясно първоначално уплътнение за поле - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Уплътняването, което да се добави към всяка страна на кутията/полето. Това бива добавяно като разширен набор от полета, използвани за премахване на фалшиви положителни резултати от маската на ИИ детектора на текст. Тези кутии са визуализирани в лилаво. - + Box Padding Extended Profile entry in Preprocessor Разширено Уплатняване на поле - + Box Right Padding Extended Profile entry in Preprocessor Разширена дясно уплътняване на полето - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Уплътняването, което да се добави към всяка страна на кутията/полето. Това се добавя към референтните полета, използвани за проба на оригиналното изображение, докато се анализира каква маска да се използва. Тези кутии са визуализирани в синьо. - + Box Reference Padding Profile entry in Preprocessor Референция на Поле - Паддинг - + Masker Profile section title Маскировчик - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Брой пиксели за разширяване на маската при всяка стъпка. Това увеличава контура на маската, така че по-малките стойности ще бъдат по-точни, но по-бавни. - + Mask Growth Step Pixels Profile entry in Masker Стъпка на растежа на маската Пиксели - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Брой стъпки за нарастване на маската. По-голямото число ще създаде повече и по-големи маски, като в крайна сметка ще бъде ограничено от размера на референтната кутия. - + Mask Growth Steps Profile entry in Masker Стъпки за растеж на маската - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Минимална дебелина на маската.о Това е като стъпка на нарастване на първата маска, а останалите ще следват стъпка на нарастване на маската Пиксели. По този начин можете да имате малък размер на стъпката, за да постигнете точност, но все пак да предотвратите много тънки маски, които могат да се използват за почистване на текст, заобиколен само от контур, когато обрисуването би било по-добрият избор. - + Min Mask Thickness Profile entry in Masker Мин. дебелина на маската - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Максимален праг на Отвъд Бяло - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker Стандартното отклонение на маска представлява вариацията на цвета по ръба на маската. За тази цел от оригиналното изображение се взема извадка само от един ред пиксели по външния ръб на маската. Ниското отклонение означава, че маската се намира в плътен цвят, което означава, че не пресича никакъв текст или други обекти. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Минимално подобрение на стандартното отклонение на маската, за да продължите да я отглеждате. Задаването на по-висока стойност тук изисква по-голямо подобрение, за да се разгледа по-голяма маска, за да се даде предимство на по-малките маски. - + Mask Improvement Threshold Profile entry in Masker Праг за подобряване на маската - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Дали да се използва алгоритъмът за бърз избор на маска. Когато е вярно, алгоритъмът за избор на маска ще избере първата перфектна маска, ако такава бъде намерена рано. Това е по-бързо, но може да не се намери най-добрата маска, ако малко по-голяма би била по-добра. - + Mask Selection Fast Profile entry in Masker Бърз избор на маска - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Обрисуване - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter ЕКСПЕРИМЕНТАЛНА ФУНКЦИЯ: Ако откриете по-добри настройки по подразбиране, моля, отворете въпрос в github, за да споделите подобренията с всички. Обърнете внимание, че инпайнтингът не е активиран по подразбиране. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter При инпайнтинга се използва машинно обучение, за да се замени съдържанието на дадено изображение въз основа на заобикалящата го среда. За маски, които не са могли да бъдат почистени добре (или изобщо), може да се използва инпайнтинг. За да се разграничи това от обезцветяването, inpainting е предназначен за значително по-лоши маски, които са имали тясно прилягане. Всички маски, които са били деноализарани, няма да бъдат инпайнтирани. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Тъй като в някои случаи тази стъпка може да доведе до лоши резултати, тя може да бъде деактивирана тук. Махнете отметката, за да деактивирате инпайнтинга. - + Inpainting Enabled Profile entry in Inpainter Обрисуване Активирано - + Inpainting Min Std Dev Profile entry in Inpainter Мин. Стд Откл за Запълване - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter Максималният радиус на маска, върху която да се извършва обрисуване. Маски, по-големи от този размер, ще бъдат оставени така, както са, защото ако маржът е толкова голям, вероятно маската вече е достатъчно добра. - + Inpainting Max Mask Radius Profile entry in Inpainter Обрисуване Маска - Макс. Радиус - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter Минималният радиус на маската, която ще се боядисва. Този параметър се добавя към оптималния размер на маската, за да се гарантира, че инпойнтът покрива цялата маска. - + Min Inpainting Radius Profile entry in Inpainter Обрисуване Маска - Мин. Радиус - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter За маски, които се оказват много по-трудни за почистване, т.е. имат високо стандартно отклонение, увеличете радиуса на обрисуване, за да покриете по-голяма част от маската. Това е допълнителен марж, който се добавя към минималния радиус на рисуване и се изчислява като: множител на радиуса на обрисуване, умножен по стандартното отклонение на маската σ - + Inpainting Radius Multiplier Profile entry in Inpainter Обрисуване Радиус‑Множител - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter Максималният радиус на маската, която ще се обрисува. Това ограничава размера, до който може да нарасне инпойнтът, за да се предотврати покриването на твърде голяма част от изображението, ако се използва голям множител на радиуса. - + Max Inpainting Radius Profile entry in Inpainter Макс. Радиус‑Обрисуване - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter След Обрисуване, изрежи от оригиналния образ, за избягване на ефекта върху други области. Това гарантира, че оригиналното изображение е запазено във възможно най-голяма степен. Радиуса е добавен към финалния контур, поради изменение само на няколко пиксели отвъд зададения регион. - + Inpainting Isolation Radius Profile entry in Inpainter Изолационен радиус на обрисуването - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker Максималното стандартно отклонение на маската, което трябва да се вземе предвид. Високата стойност тук означава по-висок толеранс за пресичане на текст или други обекти, което не е добра маска, тъй като така или иначе ще се наложи да се обрисува. Задаването на тази стойност на 0 ще позволи само перфектни маски, което се препоръчва за изображения с много висока разделителна способност. - + Mask Max Standard Deviation Profile entry in Masker Макс. Стандартно отклонение на Маска - + Color to use for the debug mask. Profile comment in Masker Цвят, използван за дебъгване на маска. - + Debug Mask Color Profile entry in Masker Цвят на дебъг-маската - + Denoiser Profile section title Неутрализатор на Шум - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser Когато е трудно да се маскира част от текста, стъпката на почистване вероятно е трябвало да избере малка маска, която оставя много артефакти от jpeg, ако те са били около текста. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser Това означава, че цветът на пикселите около ръба на маската не е равномерен, което се изразява количествено като стандартно отклонение. Неутрализатор на Шума може да се насочи към маски с минимално стандартно отклонение и да премахне областта с шум точно около маската. По този начин се запазват детайлите в останалата част от изображението, но се премахват артефактите точно около мястото, където е бил текстът. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Тъй като тази стъпка не е задължителна и може дори да е излишна за изображения с висока разделителна способност, които не страдат от артефакти на jpeg, тя може да бъде деактивирана тук. Задайте False (Невярно), за да деактивирате премахването на изолацията. - + Denoising Enabled Profile entry in Denoiser Шумо-премахване Активирано - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser Минималното стандартно отклонение на цветовете около ръба на дадена маска за извършване на шумо-премахване в областта около маската. - + Noise Min Standard Deviation Profile entry in Denoiser Шум ‑ Мин Стандартно отклонение - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Бележка: Ако обрисуването е Активирано, минималното средноквадратично отклонение на обрисуване ще действа като максимална стойност за това, след което тази маска ще бъде предадена на предходното. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser Дебелината на контура за премахване на шум около маска. - + Noise Outline Size Profile entry in Denoiser Размер на контура на шума - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Избледнявайте краищата на маската за покриване с това количество, за да прелеете плавно премахнатите части в останалата част на изображението. - + Noise Fade Radius Profile entry in Denoiser Радиус на затихване на шума - + Set to True to support colored images. Profile comment in Denoiser Задай на Вярно (True) за поддръжка на цветни изображения. - + Colored Images Profile entry in Denoiser Цветни Изображения - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser В каква степен да се приложи шумо‑редуциране на образа. По-високи стойности значат повече шумо-премахване. - + Filter Strength Profile entry in Denoiser Сила на Филтъра - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser Колко силно да се премахнат цветовите канали, за сметка на яркостта. По-високите стойности означават по-силно понижаване на шума. - + Color Filter Strength Profile entry in Denoiser Сила на цветния филтър - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Размер в пиксели на шаблонното петно, което се използва за изчисляване на тежестта. Трябва да е нечетен. - + Template Window Size Profile entry in Denoiser Размер на прозореца на шаблона - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Размер в пиксели на прозореца, използван при изчисляване на осреднена тежест на даден пиксел. Трябва да е нечетно. - + Search Window Size Profile entry in Denoiser Големина на прозореца за търсене @@ -4249,7 +4254,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4666,14 +4671,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>Пътят на модела на детектора на текст {value} не съществува, връща се към настройките по подразбиране. Можете да изтеглите ръчно оттук <a href="https://github.com/zyddnys/manga-image-translator/releases/latest"></a> или продължете да използвате стандартния модел.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From 0cc5e4096fb056481cd85dfbbe393cd5207ac4c3 Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:12 +0100 Subject: [PATCH 08/18] New translations panelcleaner_source.ts (Czech) --- translations/PanelCleaner_cs_CZ.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_cs_CZ.ts b/translations/PanelCleaner_cs_CZ.ts index ff1d50c..1062ea1 100644 --- a/translations/PanelCleaner_cs_CZ.ts +++ b/translations/PanelCleaner_cs_CZ.ts @@ -4,7 +4,7 @@ - + default reserved profile name výchozí @@ -20,54 +20,54 @@ Profil {profile_name} vytvořen. - + Profile name cannot be empty. Název profilu nemůže být prázdný. - + Profile name already in use. Název profilu je již použit. - + Profile name is reserved. Název profilu je vyhrazen. - - - - + + + + Unknown Neznámý - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Stupnice šedi - + 1-bit Color mode 1 bit - + Palette Color mode Paleta @@ -78,45 +78,45 @@ Chyba při stahování souboru z url: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Chyba při stahování souboru z url: {url} Nepodařilo se uložit soubor do {save_path} - + Error downloading file from url: {url} The file content is different from expected. Chyba při stahování souboru z url: {url} Obsah souboru je jiný, než se očekávalo. + - - - + + + - - - + + N/A N/A - + OCR Analytics Analýza OCR - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Počet políček: {num_boxes} | Počet malých políček: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Počet odstraněných políček: {num_removed} ({removed_box_ratio} celkem, {removed_among_small_ratio} malých políček) @@ -155,66 +155,66 @@ Odstraněné bubliny: - + Small boxes Malé políčka - + Removed boxes Odstraněné políčka - - + + Failed Neúspěšné - + Perfect Perfektní - + Total Celkem - + Mask Fitment Analytics Analýza uložení masky - + Total Celkem - + Total boxes Celkem políček - + Masks succeeded Úspěšné masky - + Masks failed Neúspěšné masky - + Perfect masks Perfektní masky - + Average border deviation Průměrná odchylka okraje @@ -233,7 +233,7 @@ Použití masky podle tloušťky (v pixelech): Maska ({thickness}px) - + Box mask Maska políček @@ -254,27 +254,27 @@ Stránky s chybami / celkem: - + Denoising Analytics Analýza odstranění šumu - + Total masks Celkem masek - + Masks denoised Masky s odstraněním šumu - + Minimum deviation to denoise Minimální odchylka pro odstranění šumu - + Maximum allowed deviation Maximální povolená odchylka @@ -293,27 +293,27 @@ Stránky s chybami / celkem: Šum odstraněn - + Inpainting Analytics Analýza retušování - + Inpainting performed Provedené retušování - + Average thickness Průměrná tloušťka - + Minimum thickness Minimální tloušťka - + Maximum thickness Maximální tloušťka @@ -372,7 +372,7 @@ Stránky s chybami / celkem: Formát souboru nebyl rozpoznán. - + The CSV file must start with a header row, followed by data rows. Soubor CSV musí začínat řádkem záhlaví, po kterém následují řádky s daty. @@ -402,7 +402,7 @@ Stránky s chybami / celkem: Cestě k souboru odpovídá více obrázkům: {path} - + Error, no version found. Chyba, nebyla nalezena žádná verze. @@ -412,7 +412,7 @@ Stránky s chybami / celkem: Chyba CUDA - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Váše GPU nepodporuje požadované operace CUDA.<br><br>Zkuste odinstalovat aktuální verzi torch a torchvision a místo nich nainstalovat verzi pro CPU (nebo jinou verzi CUDA).<br>Další pokyny najdete zde: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Podívejte se do sekce "Compute Platform", kde najdete dostupné verze.<br><br>Vaše aktuální verze CUDA je: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Stránky s chybami / celkem: Nebyl nalezen model pro retušování. - + filename název souboru - + startx startx - + starty starty - + endx endx - + endy endy - + text text - + Saved detected text to {output_file} Detekovaný text byl uložen do {output_file} - + Failed to write detected text to {output_file} Nepodařilo se zapsat detekovaný text do {output_file} - + Save Failed Ukládání selhalo - + Failed to write detected text to file. Nepodařilo se zapsat detekovaný text do souboru. @@ -546,7 +546,7 @@ Stránky s chybami / celkem: - + Page layered export @@ -635,7 +635,7 @@ Nástroj s umělou inteligencí na čištění panelů mangy. ColorButton - + Select Color Vybrat barvu @@ -681,7 +681,7 @@ Nástroj s umělou inteligencí na čištění panelů mangy. Nepodporovaný systém - + This system is not supported. Tento systém není podporován. @@ -696,12 +696,12 @@ Nástroj s umělou inteligencí na čištění panelů mangy. Odinstalovat z Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. Rozšíření bude nainstalováno do {path}. Nevyžaduje oprávnění root. - + The extension file will be deleted. This doesn't require root permissions. Rozšíření bude odstraněno. Nevyžaduje oprávnění root. @@ -716,17 +716,17 @@ Nástroj s umělou inteligencí na čištění panelů mangy. Odinstalovat z Průzkumníka souborů - + The extension will be installed in the Windows registry.This requires admin permissions. Rozšíření se nainstaluje do registru Windows. Vyžaduje oprávnění správce. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. Rozšíření bude odinstalováno z registru Windows. Vyžaduje oprávnění správce. - + Dolphin extension installed successfully. Rozšíření Dolphin bylo úspěšně nainstalováno. @@ -736,7 +736,7 @@ Nástroj s umělou inteligencí na čištění panelů mangy. Instalace proběhla úspěšně - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Rozšíření Dolphin bylo úspěšně nainstalováno. @@ -816,12 +816,12 @@ Ve výchozím nastavení se používá výchozí profil. Toto výchozí nastaven FileTable - + Original size in pixels (width × height) Původní velikost v pixelech (šířka × výška) - + Processing size in pixels (width × height), scale factor Velikost zpracování v pixelech (šířka × výška), faktor měřítka @@ -847,7 +847,7 @@ Ve výchozím nastavení se používá výchozí profil. Toto výchozí nastaven Nepodporované TIFF soubory - + The following 5-channel TIFF files are not supported: Následující pětikanálové soubory TIFF nejsou podporovány: @@ -874,27 +874,27 @@ Ve výchozím nastavení se používá výchozí profil. Toto výchozí nastaven rozdělit - + Number of boxes removed by the OCR model / total boxes Počet políček odstraněných modelem OCR / celkový počet políček - + Number of boxes that failed to generate a mask / total boxes Počet políček, u kterých se nepodařilo vygenerovat masku / celkový počet políček - + Number of boxes that were perfectly masked / total boxes Počet políček, které byly dokonale zamaskovány / celkový počet políček - + Number of boxes that were denoised / total boxes Počet políček, které byly zbaveny šumu / celkový počet políček - + Number of boxes that were inpainted Počet políček, který byly retušovány @@ -1042,14 +1042,14 @@ Ve výchozím nastavení se používá výchozí profil. Toto výchozí nastaven Nemožná žádost - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Odstranění šumu je zakázáno v profilu, výstup nelze vygenerovat. Povolte prosím odstranění šumu v profilu a zkuste to znovu. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Retušování je zakázáno v profilu, výstup nelze vygenerovat. @@ -1071,7 +1071,7 @@ Povolte prosím retušování v profilu a zkuste to znovu. Export se nezdařil - + Failed to export image: Nepodařilo se exportovat obrázek: @@ -1114,7 +1114,7 @@ Povolte prosím retušování v profilu a zkuste to znovu. výsledek - + results výsledky @@ -1129,17 +1129,17 @@ Povolte prosím retušování v profilu a zkuste to znovu. nový obrázek - + new images nové obrázky - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Odpovídá {num_images} {plural_images} výsledkům OCR. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} osiřelé OCR {plural_result} budou ztraceny. @@ -1238,12 +1238,12 @@ Povolte prosím retušování v profilu a zkuste to znovu. Poznámka: Jméno "{name}" bylo skryto - + Corrupted log session Poškozený protokol - + Current session Tato relace @@ -1258,7 +1258,7 @@ Povolte prosím retušování v profilu a zkuste to znovu. Chyba - + Errors Chyby @@ -1268,7 +1268,7 @@ Povolte prosím retušování v profilu a zkuste to znovu. Kritická - + Criticals Kritické @@ -1284,49 +1284,49 @@ Povolte prosím retušování v profilu a zkuste to znovu. MainWindow - + Open Config Statusbar button Otevřít konfiguraci - + Open Log Statusbar button Otevřít protokol - + CUDA Enabled CUDA k dispozici - + Select Output Directory Vybrat výstupní adresář - + Aborting... Rušení... - + Loading OCR model... Načítání modelu OCR... - + Loaded OCR model. OCR model načten. - + Error Chyba - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1335,7 +1335,7 @@ Continue anyway? Přesto pokračovat? - + Delete Models Smazat modely @@ -1345,23 +1345,23 @@ Přesto pokračovat? Jste si jisti, že chcete odstranit modely strojového učení? Znemožníte tím čištění a OCR, dokud nebudou znovu staženy. - + Encountered error: Vyskytla se chyba: - + MPS Enabled MPS povoleno - - + + Inpainting Model Update Aktualizace modelu pro retušování - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1375,148 +1375,148 @@ Přejít na nový model? Starý model byl zachován. Chcete-li provést aktualizaci, vyberte v nabídce "Nápověda" a poté "Odstranit modely strojového učení". - + Layout Reset Obnovit rozložení - + The window layout has been reset. The default layout will be used next time. Rozložení okna bylo resetováno. Příště bude použito výchozí rozložení. - + Multiple Instances Více instancí - + Found process ID in lock file: Nalezeno ID procesu v zamčeném souboru: - + Failed to delete models. Nepodařilo se odstranit modely. - + Delete Failed Odstranění se nezdařilo - + Models Deleted Modely odstraněny - + The models were deleted. Would you like to download them again? Modely byly odstraněny. Chcete si je znovu stáhnout? - + Not Supported Není podporováno - + Unsupported system or package format Nepodporovaný formát systému nebo balíčku - + Shutdown Vypnout - + After processing, the system will shut down. Po dokončení se systém vypne. - + You have review options enabled, these will need to be manually closed before the action can start. Máte povolenou možnost přezkoumání, které je třeba před spuštěním akce ručně zavřít. - + You have disabled writing output. Zakázali jste zápis výstupu. - + Shutting down system... Vypnání systému... - + Running post action "{action}"... Spuštění post-akce "{action}"... - + Restart Required Požaduje se restart - + The language has been changed. Please restart the application for the changes to take effect. The language has been changed. Please restart the application for the changes to take effect. - - + + System Language System Language - + Import Profile Import Profile - + Profile Files (*.conf) Profile Files (*.conf) - + Profile Imported Profile Imported - + Import Error Import Error - + Failed to Delete Failed to Delete - + The default profile cannot be deleted. The default profile cannot be deleted. - + Delete Profile Delete Profile - + Are you sure you want to delete the profile {profile_name}? Are you sure you want to delete the profile {profile_name}? - + Delete Error Delete Error - + Failed to delete the profile. Failed to delete the profile. @@ -1528,217 +1528,217 @@ Switching profiles will discard changes to the current profile. Switching profiles will discard changes to the current profile. - + Unsaved changes Unsaved changes - - + + Load Error Load Error - - + + Failed to load profile: Failed to load profile: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Profile Created - + Create Error Create Error - + Failed to save profile to {profile_path} Failed to save profile to {profile_path} - - + + Save Error Save Error - + Failed to save profile. Failed to save profile. - + Profile saved to {profile_path} Profile saved to {profile_path} - + Failed to save config. Failed to save config. - + Failed to save the new profile to the configuration file. Failed to save the new profile to the configuration file. - - + + No Files No Files - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. - + No Outputs No Outputs - + No outputs were requested. Please select at least one output before cleaning. No outputs were requested. Please select at least one output before cleaning. - - + + cleaned cleaned - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists File Exists - + The file '{output_path}' already exists. Overwrite? The file '{output_path}' already exists. Overwrite? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Processing Finished - + Finished processing all files. Finished processing all files. - + Processing Aborted Processing Aborted - + Processing aborted. Processing aborted. - + Processing Error Processing Error - + Encountered an error while processing files. Encountered an error while processing files. @@ -1749,7 +1749,7 @@ Check the online documentation for how to add support for more languages. - + Profile Profile @@ -1764,360 +1764,365 @@ Check the online documentation for how to add support for more languages.Apply - - + + Save Save - + Reset All Reset All - + Drag and Drop Images or Folders Here Drag and Drop Images or Folders Here - + Images Images - - + + File File - + Size Size - + Processing Size The size that the picture is shrunk down to for processing. Processing Size - + File Size File Size - + Color Mode Like RGB, grayscale etc. Color Mode - + Analytics Can also call this statistics. Analytics - + Process Process - + Cleaning Cleaning - + OCR OCR - + Edit Existing Output Edit Existing Output - + Abort Abort - + Start Start - + Output Output - + Cleaned Image Cleaned Image - + Mask Mask - + Isolated Text Isolated Text - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Save Output - + Output Directory: Output Directory: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. You can use a relative path to create a subfolder at the image's original location, or use an absolute path. - + Plain Text Plain Text - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Output File: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Generating: - + Current Step: Current Step: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Images Processed: - + Set Default Set Default - + Settings Settings - + Theme As in color theme Theme - + Language Language - + Help Help - + Add Files... Add Files... - + Remove All Files Remove All Files - + Remove File Remove File - + Help Translate Panel Cleaner Help Translate Panel Cleaner - + Report an Issue... Report an Issue... - + Simulate Exception Simulate Exception - + Add File Browser Extension... Add File Browser Extension... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New New - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Delete - + Import... Import... - + Online Documentation Online Documentation - + About Panel Cleaner About Panel Cleaner - + View License View License - + Save as... Save as... - + System System - + Dark Dark - + Light Light - + Add Folder... Add Folder... - + Delete Machine Learning Models Delete Machine Learning Models - + Download Machine Learning Models Download Machine Learning Models - + Donate Donate @@ -2149,19 +2154,19 @@ Check the online documentation for how to add support for more languages.OCR model: - + Already downloaded Already downloaded - + Text Detector model (CUDA) Text Detector model (CUDA) - + Text Detector model (CPU) Text Detector model (CPU) @@ -2180,7 +2185,7 @@ Check the online documentation for how to add support for more languages.OCR download errors - + ETA estimated time of completion ETA @@ -2758,7 +2763,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2839,73 +2844,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input The original image with scaling applied (if needed). - + Input Step name in the image details view Input - + Input Output name in the image details view Input - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask The rough mask generated by the AI. - + Text Detection Step name in the image details view Text Detection - + AI Mask Output name in the image details view AI Mask - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes The unfiltered box data generated by the AI. - + Raw Boxes Output name in the image details view Raw Boxes - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes The outlines of the text boxes the AI found. - + Preprocessor Step name in the image details view Preprocessor - + Initial Boxes Output name in the image details view Initial Boxes - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2913,181 +2918,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. - + Final Boxes Output name in the image details view Final Boxes - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask The mask of the merged boxes. - + Masker Step name in the image details view Masker - + Box Mask Output name in the image details view Box Mask - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask The rough text detection mask with everything outside the box mask cut out. - + Cut Mask Output name in the image details view Cut Mask - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers The different steps of growth around the cut mask displayed in different colors. - + Mask Layers Output name in the image details view Mask Layers - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). - + Fitment Quality Output name in the image details view Fitment Quality - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay The input image with the final mask overlaid in color. - + Mask Overlay Output name in the image details view Mask Overlay - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask The collection of masks for each bubble that fit best. - + Final Mask Output name in the image details view Final Mask - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text The text layer isolated from the input image. - + Isolated Text Output name in the image details view Isolated Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output The input image with the final mask applied. - + Masked Output Output name in the image details view Masked Output - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask The masks that required denoising, to be overlaid on the final mask when exporting. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Denoise Mask - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output The input image with the denoised mask applied. - + Denoised Output Output name in the image details view Denoised Output - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask The inpainted sections around the text that was poorly cleaned, if at all. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Inpainted Mask - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output The input image: cleaned, denoised (if enabled), and inpainted. - + Inpainted Output Output name in the image details view Inpainted Output - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3096,1141 +3101,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title General - + About this profile: Profile comment in General About this profile: - + Notes Profile entry in General Notes - + Preferred file type to save the cleaned image as. Profile comment in General Preferred file type to save the cleaned image as. - + Preferred File Type Profile entry in General Preferred File Type - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Preferred file type to save the mask as. Only image formats that allow for transparency are supported. - + Preferred Mask File Type Profile entry in General Preferred Mask File Type - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. - + Input Height Lower Target Profile entry in General Input Height Lower Target - + Input Height Upper Target Profile entry in General Input Height Upper Target - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Text Detector - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> - + Model Path Profile entry in Text Detector Model Path - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! - + Concurrent Models Profile entry in Text Detector Concurrent Models - + Preprocessor Profile section title Preprocessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Minimum size of any box to keep it. - + Box Min Size Profile entry in Preprocessor Box Min Size - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. - + Suspicious Box Min Size Profile entry in Preprocessor Suspicious Box Min Size - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. - + Box Overlap Threshold Profile entry in Preprocessor Box Overlap Threshold - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. - + OCR Enabled Profile entry in Preprocessor OCR Enabled - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. - + OCR Use Tesseract Profile entry in Preprocessor OCR Use Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. - + OCR Engine Profile entry in Preprocessor OCR Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. - + Reading Order Profile entry in Preprocessor Reading Order - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. - + OCR Max Size Profile entry in Preprocessor OCR Max Size - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR Blacklist Pattern - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor OCR Strict Language - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. - + Box Padding Initial Profile entry in Preprocessor Box Padding Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. - + Box Right Padding Initial Profile entry in Preprocessor Box Right Padding Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Box Right Padding Extended - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. - + Box Reference Padding Profile entry in Preprocessor Box Reference Padding - + Masker Profile section title Masker - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. - + Mask Growth Step Pixels Profile entry in Masker Mask Growth Step Pixels - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. - + Mask Growth Steps Profile entry in Masker Mask Growth Steps - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. - + Min Mask Thickness Profile entry in Masker Min Mask Thickness - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Off White Max Threshold - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. - + Mask Improvement Threshold Profile entry in Masker Mask Improvement Threshold - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. - + Mask Selection Fast Profile entry in Masker Mask Selection Fast - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting Enabled - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. - + Min Inpainting Radius Profile entry in Inpainter Min Inpainting Radius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation - + Inpainting Radius Multiplier Profile entry in Inpainter Inpainting Radius Multiplier - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. - + Max Inpainting Radius Profile entry in Inpainter Max Inpainting Radius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. - + Inpainting Isolation Radius Profile entry in Inpainter Inpainting Isolation Radius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. - + Mask Max Standard Deviation Profile entry in Masker Mask Max Standard Deviation - + Color to use for the debug mask. Profile comment in Masker Color to use for the debug mask. - + Debug Mask Color Profile entry in Masker Debug Mask Color - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. - + Denoising Enabled Profile entry in Denoiser Denoising Enabled - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. - + Noise Min Standard Deviation Profile entry in Denoiser Noise Min Standard Deviation - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser The thickness of an outline to denoise around a mask. - + Noise Outline Size Profile entry in Denoiser Noise Outline Size - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. - + Noise Fade Radius Profile entry in Denoiser Noise Fade Radius - + Set to True to support colored images. Profile comment in Denoiser Set to True to support colored images. - + Colored Images Profile entry in Denoiser Colored Images - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the image. Higher values mean more denoising. - + Filter Strength Profile entry in Denoiser Filter Strength - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. - + Color Filter Strength Profile entry in Denoiser Color Filter Strength - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Size in pixels of the template patch that is used to compute weights. Should be odd. - + Template Window Size Profile entry in Denoiser Template Window Size - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. - + Search Window Size Profile entry in Denoiser Search Window Size @@ -4257,7 +4262,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4674,14 +4679,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From 05903497660dce5e42307dafc66deb55673c8ec8 Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:13 +0100 Subject: [PATCH 09/18] New translations panelcleaner_source.ts (German) --- translations/PanelCleaner_de_DE.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_de_DE.ts b/translations/PanelCleaner_de_DE.ts index b5a8b85..2d82067 100644 --- a/translations/PanelCleaner_de_DE.ts +++ b/translations/PanelCleaner_de_DE.ts @@ -4,7 +4,7 @@ - + default reserved profile name standard @@ -20,54 +20,54 @@ Profil {profile_name} erstellt. - + Profile name cannot be empty. Der Profilname darf nicht leer sein. - + Profile name already in use. Der Profilname wird bereits verwendet. - + Profile name is reserved. Der Profilname ist reserviert. - - - - + + + + Unknown Unbekannt - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Graustufen - + 1-bit Color mode 1-Bit - + Palette Color mode Palette @@ -78,45 +78,45 @@ Fehler beim Herunterladen einer Datei von: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Fehler beim Herunterladen der Datei von: {url} Die Datei konnte nicht in {save_path} gespeichert werden. - + Error downloading file from url: {url} The file content is different from expected. Fehler beim Herunterladen einer Datei von: {url} Der Inhalt der Datei ist anders als erwartet. + - - - + + + - - - + + N/A K.A. - + OCR Analytics OCR Statistik - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Felderanzahl: {num_boxes} | Anzahl kleinen Felder: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Anzahl entfernter Felder: {num_removed} ({removed_box_ratio} insgesamt, {removed_among_small_ratio} kleiner Felder) @@ -155,66 +155,66 @@ Entfernte Blasen: - + Small boxes Kleine Felder - + Removed boxes Entfernte Felder - - + + Failed Gescheitert - + Perfect Perfekt - + Total Insgesamt - + Mask Fitment Analytics Maskenanpassungsstatistik - + Total Insgesamt - + Total boxes Felder insgesamt - + Masks succeeded Masken erfolgreich - + Masks failed Masken gescheitert - + Perfect masks Perfekte Masken - + Average border deviation Durchschn. Grenzabweichung @@ -233,7 +233,7 @@ Maskenverwendung nach Dicke (in Pixel): Maske ({thickness}px) - + Box mask Felder-Maske @@ -254,27 +254,27 @@ Seiten mit Fehlern / insgesamt: - + Denoising Analytics Entrauschungsstatistik - + Total masks Masken insgesamt - + Masks denoised Masken entrauscht - + Minimum deviation to denoise Min. Abweichung zum Entrauschen - + Maximum allowed deviation Max. zulässige Abweichung @@ -293,27 +293,27 @@ Seiten mit Fehlern / insgesamt: Entrauscht - + Inpainting Analytics Übermalungsstatistik - + Inpainting performed Übermalung durchgeführt - + Average thickness Durchschnittliche Dicke - + Minimum thickness Minimale Dicke - + Maximum thickness Maximale Dicke @@ -372,7 +372,7 @@ Seiten mit Fehlern / insgesamt: Das Dateiformat wurde nicht erkannt. - + The CSV file must start with a header row, followed by data rows. Die CSV-Datei muss mit einer Kopfzeile beginnen, gefolgt von Datenzeilen. @@ -402,7 +402,7 @@ Seiten mit Fehlern / insgesamt: Mehrere Bilder stimmen mit dem Dateipfad überein: {path} - + Error, no version found. Fehler, keine Version gefunden. @@ -412,7 +412,7 @@ Seiten mit Fehlern / insgesamt: CUDA-Fehler - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Ihr Grafikprozessor unterstützt die erforderlichen CUDA-Operationen nicht.<br><br>Deinstallieren Sie die aktuellen Versionen von torch und torchvision und installieren Sie stattdessen die CPU-Version (oder eine andere CUDA-Version).<br>Weitere Anweisungen finden Sie hier: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Überprüfen Sie den Abschnitt "Compute Platform", um die verfügbaren Versionen zu sehen.<br><br>Ihre aktuelle CUDA-Version ist: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Seiten mit Fehlern / insgesamt: Übermalungsmodell nicht gefunden. - + filename Dateiname - + startx startx - + starty starty - + endx endx - + endy endy - + text Text - + Saved detected text to {output_file} Erkannter Text wurde in {output_file} gespeichert. - + Failed to write detected text to {output_file} Der erkannte Text konnte nicht in {output_file} geschrieben werden. - + Save Failed Speichern Fehlgeschlagen - + Failed to write detected text to file. Der erkannte Text konnte nicht in die Datei geschrieben werden. @@ -538,7 +538,7 @@ Seiten mit Fehlern / insgesamt: - + Page layered export @@ -625,7 +625,7 @@ Reinige Mangas durch KI. ColorButton - + Select Color Farbe Auswählen @@ -671,7 +671,7 @@ Reinige Mangas durch KI. Nicht unterstütztes System - + This system is not supported. Dieses System wird nicht unterstützt. @@ -686,12 +686,12 @@ Reinige Mangas durch KI. Deinstallation für Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. Die Erweiterung wird unter {path} installiert. Dies erfordert keine Root-Rechte. - + The extension file will be deleted. This doesn't require root permissions. Die Erweiterungsdatei wird gelöscht. Dies erfordert keine Root-Rechte. @@ -706,17 +706,17 @@ Reinige Mangas durch KI. Deinstallation für Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. Die Erweiterung wird in der Windows-Registry installiert. Dies erfordert Admin-Rechte. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. Die Erweiterung wird aus der Windows-Registry deinstalliert. Dies erfordert Admin-Rechte. - + Dolphin extension installed successfully. Dolphin-Erweiterung erfolgreich installiert. @@ -726,7 +726,7 @@ Reinige Mangas durch KI. Installation erfolgreich - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Dolphin-Erweiterung erfolgreich installiert. @@ -806,12 +806,12 @@ Standardmäßig wird das eingebaute Profil verwendet. Sie können diese Standard FileTable - + Original size in pixels (width × height) Originalgröße in Pixel (Breite ⋅ Höhe) - + Processing size in pixels (width × height), scale factor Verarbeitungsgröße in Pixel (Breite ⋅ Höhe), Skalierungsfaktor @@ -837,7 +837,7 @@ Standardmäßig wird das eingebaute Profil verwendet. Sie können diese Standard Nicht unterstützte TIFF-Dateien - + The following 5-channel TIFF files are not supported: Die folgenden 5-Kanal-TIFF-Dateien werden nicht unterstützt: @@ -864,27 +864,27 @@ Standardmäßig wird das eingebaute Profil verwendet. Sie können diese Standard abschnitt - + Number of boxes removed by the OCR model / total boxes Anzahl der durch das OCR-Modell entfernten Felder / Felder insgesamt - + Number of boxes that failed to generate a mask / total boxes Anzahl der Felder, für die keine Maske erzeugt werden konnte / Felder insgesamt - + Number of boxes that were perfectly masked / total boxes Anzahl der perfekt maskierten Felder / Felder insgesamt - + Number of boxes that were denoised / total boxes Anzahl der entrauschten Felder / Felder insgesamt - + Number of boxes that were inpainted Anzahl der Felder, die übermalt wurden @@ -1032,14 +1032,14 @@ Standardmäßig wird das eingebaute Profil verwendet. Sie können diese Standard Unmöglicher Auftrag - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Die Entrauschung ist im aktuellen Profil deaktiviert, diese Ausgabe kann nicht erzeugt werden. Bitte aktivieren Sie die Entrauschung in den Profileinstellungen und versuchen Sie es erneut. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Übermalung ist im aktuellen Profil deaktiviert, diese Ausgabe kann nicht erzeugt werden. @@ -1061,7 +1061,7 @@ Bitte aktivieren Sie Übermalung in den Profileinstellungen und versuchen Sie es Export fehlgeschlagen - + Failed to export image: Bild kann nicht exportiert werden: @@ -1104,7 +1104,7 @@ Bitte aktivieren Sie Übermalung in den Profileinstellungen und versuchen Sie es Ergebnis - + results Ergebnisse @@ -1119,17 +1119,17 @@ Bitte aktivieren Sie Übermalung in den Profileinstellungen und versuchen Sie es neues Bild - + new images neue Bilder - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Zugewiesen: {num_images} {plural_images} mit OCR-Ergebnissen. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} verwaiste OCR {plural_result} geht verloren. @@ -1228,12 +1228,12 @@ Bitte aktivieren Sie Übermalung in den Profileinstellungen und versuchen Sie es Hinweis: Der Name "{name}" wurde ausgeblendet. - + Corrupted log session Protokoll unlesbar - + Current session Aktuelle Sitzung @@ -1248,7 +1248,7 @@ Bitte aktivieren Sie Übermalung in den Profileinstellungen und versuchen Sie es Fehler - + Errors Fehler @@ -1258,7 +1258,7 @@ Bitte aktivieren Sie Übermalung in den Profileinstellungen und versuchen Sie es Kritisch - + Criticals Kritisch @@ -1274,49 +1274,49 @@ Bitte aktivieren Sie Übermalung in den Profileinstellungen und versuchen Sie es MainWindow - + Open Config Statusbar button Konfiguration öffnen - + Open Log Statusbar button Log öffnen - + CUDA Enabled CUDA Aktiviert - + Select Output Directory Ausgabeverzeichnis Auswählen - + Aborting... Abbrechen... - + Loading OCR model... OCR-Modell lädt... - + Loaded OCR model. OCR-Modell geladen. - + Error Fehler - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1325,7 +1325,7 @@ Continue anyway? Trotzdem fortfahren? - + Delete Models Modelle Löschen @@ -1335,23 +1335,23 @@ Trotzdem fortfahren? Sind Sie sicher, dass Sie die Modelle für maschinelles Lernen löschen wollen? Dadurch werden Reinigung und OCR unmöglich, bis sie erneut heruntergeladen werden. - + Encountered error: Ein Fehler ist aufgetreten: - + MPS Enabled MPS Aktiviert - - + + Inpainting Model Update Übermal-Modell Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1365,148 +1365,148 @@ Auf das neue Modell umsteigen? Das alte Modell bleibt erhalten. Wählen Sie zum Aktualisieren "Hilfe" und dann "Modelle für maschinelles Lernen löschen" in der Menüleiste. - + Layout Reset Layout zurücksetzen - + The window layout has been reset. The default layout will be used next time. Das Fensterlayout wurde zurückgesetzt. Beim nächsten Mal wird das Standardlayout verwendet. - + Multiple Instances Mehrere Instanzen - + Found process ID in lock file: Prozess-ID in Sperrdatei gefunden: - + Failed to delete models. Modelldaten konnten nicht gelöscht werden. - + Delete Failed Löschen Fehlgeschlagen - + Models Deleted Modelldaten Gelöscht - + The models were deleted. Would you like to download them again? Die Modelldaten wurden gelöscht. Möchten Sie sie erneut herunterladen? - + Not Supported Nicht Unterstützt - + Unsupported system or package format Dieses System oder diese Installationsmethode sind nicht unterstützt. - + Shutdown Herunterfahren - + After processing, the system will shut down. Nach der Bearbeitung wird das System heruntergefahren. - + You have review options enabled, these will need to be manually closed before the action can start. Sie haben Überprüfungsoptionen aktiviert. Diese müssen manuell geschlossen werden, bevor die Aktion gestartet werden kann. - + You have disabled writing output. Sie haben die Schreibausgabe deaktiviert. - + Shutting down system... Wird heruntergefahren... - + Running post action "{action}"... Nachlaufaktion "{action}" wird ausgeführt... - + Restart Required Neustart Erforderlich - + The language has been changed. Please restart the application for the changes to take effect. Die Sprache wurde geändert. Bitte starten Sie die Anwendung neu, damit die Änderungen wirksam werden. - - + + System Language Systemsprache - + Import Profile Profil Importieren - + Profile Files (*.conf) Profildateien (*.conf) - + Profile Imported Profil Importiert - + Import Error Fehler beim Importieren - + Failed to Delete Löschen Fehlgeschlagen - + The default profile cannot be deleted. Das Standardprofil kann nicht gelöscht werden. - + Delete Profile Profil Löschen - + Are you sure you want to delete the profile {profile_name}? Sind Sie sicher, dass Sie das Profil {profile_name} löschen möchten? - + Delete Error Löschen Fehlgeschlagen - + Failed to delete the profile. Das Profil konnte nicht gelöscht werden. @@ -1518,217 +1518,217 @@ Switching profiles will discard changes to the current profile. Wenn Sie das Profil wechseln, werden die Änderungen am aktuellen Profil verworfen. - + Unsaved changes Ungespeicherte Änderungen - - + + Load Error Fehler beim Laden - - + + Failed to load profile: Das Profil konnte nicht geladen werden: - + Tesseract OCR is not installed or not found Tesseract OCR ist nicht installiert oder wird nicht gefunden - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <'html>Tesseract konnte nicht für OCR ausgeführt werden. Manga-ocr wird wieder verwendet. Bitte lesen Sie die Anweisungen zur korrekten Installation von Tesseract <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">hier</a> oder verwenden Sie weiterhin das Standardmodell.</html> - + Unsupported Language Nicht unterstützte Sprache - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. Die Sprache '{lang_name}' wird von keiner Ihrer aktuellen OCR-Modelle unterstützt. Lesen Sie in der Online-Dokumentation nach, wie Sie die Unterstützung für weitere Sprachen hinzufügen können. - + Profile Created Profil Erstellt - + Create Error Fehler beim Erstellen - + Failed to save profile to {profile_path} Das Profil konnte nicht in {profile_path} gespeichert werden - - + + Save Error Fehler beim Speichern - + Failed to save profile. Das Profil konnte nicht gespeichert werden. - + Profile saved to {profile_path} Profil gespeichert in {profile_path} - + Failed to save config. Konfiguration konnte nicht gespeichert werden. - + Failed to save the new profile to the configuration file. Das neue Profil konnte nicht in der Konfigurationsdatei gespeichert werden. - - + + No Files Keine Dateien - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. Keine Dateien zu verarbeiten. Sie können Dateien in der Mitte des Fensters ablegen oder über die Menüleiste hinzufügen: Dateien -> Dateien hinzufügen oder Ordner hinzufügen. - + No Outputs Keine Ausgaben - + No outputs were requested. Please select at least one output before cleaning. Es wurden keine Ausgaben angefordert. Bitte wählen Sie vor der Reinigung mindestens eine Ausgabe aus. - - + + cleaned gereinigt - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. Keine Dateien zu verarbeiten. Um eine alte OCR-Ausgabedatei zu bearbeiten, müssen Sie zunächst (eines oder mehrere) der Bilder laden, zu denen sie gehört. - + Open OCR Output File OCR-Ausgabedatei öffnen - + OCR Output Files (*.txt *.csv) OCR-Ausgabedateien (*.txt *.csv) - + Parse Error Parse-Fehler - + Failed to parse the OCR output file. Die OCR-Ausgabedatei konnte nicht gelesen werden. - + File Exists Datei Bereits Vorhanden - + The file '{output_path}' already exists. Overwrite? Die Datei '{output_path}' existiert bereits. Wollen Sie diese überschreiben? - + Export Images Bilder exportieren - + Would you like to export the cleaned images? Möchten Sie die bereinigten Bilder exportieren? - + Export OCR Results OCR-Ergebnisse Exportieren - + After processing, the "{action}" action will be executed. Nach der Bearbeitung wird die Aktion "{action}" ausgeführt. - - + + Post action canceled. Nachlaufaktion abgebrochen. - + Un-Applied Changes Nicht angewandte Änderungen - + You have un-applied changes to the profile. Continue anyway? Sie haben die Änderungen am Profil noch nicht übernommen. Trotzdem fortfahren? - + Would you like to export the OCR results? Möchten Sie die OCR-Ergebnisse exportieren? - + Processing Finished Verarbeitung Beendet - + Finished processing all files. Die Verarbeitung aller Dateien ist abgeschlossen. - + Processing Aborted Abbruch - + Processing aborted. Die Verarbeitung wurde abgebrochen. - + Processing Error Verarbeitungsfehler - + Encountered an error while processing files. Bei der Verarbeitung von Dateien ist ein Fehler aufgetreten. @@ -1739,7 +1739,7 @@ Lesen Sie in der Online-Dokumentation nach, wie Sie die Unterstützung für weit - + Profile Profil @@ -1754,360 +1754,365 @@ Lesen Sie in der Online-Dokumentation nach, wie Sie die Unterstützung für weit Anwenden - - + + Save Speichern - + Reset All Alle zurücksetzen - + Drag and Drop Images or Folders Here Bilder oder Ordner hier ablegen - + Images Bilder - - + + File Datei - + Size Größe - + Processing Size The size that the picture is shrunk down to for processing. Verarbeitungsgröße - + File Size Dateigröße - + Color Mode Like RGB, grayscale etc. Farbmodus - + Analytics Can also call this statistics. Statistik - + Process Prozess - + Cleaning Reinigung - + OCR OCR - + Edit Existing Output Vorhandene Ausgabe bearbeiten - + Abort Abbrechen - + Start Starten - + Output Ausgabe - + Cleaned Image Gereinigtes Bild - + Mask Maske - + Isolated Text Isolierter Text - - + + Review Output Überprüfung der Ausgabe - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>Wenn dieses Kontrollkästchen aktiviert ist, wird nach Abschluss des Vorgangs ein Überprüfungsfenster geöffnet. Die Ergebnisse können auch in der Detailansicht jedes Bildes eingesehen werden, bevor sie auf der Festplatte gespeichert werden.</p></body></html> - + Save Output Ausgabe speichern - + Output Directory: Ausgabeverzeichnis: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. Sie können einen relativen Pfad verwenden, um einen Unterordner am ursprünglichen Speicherort des Bildes zu erstellen, oder einen absoluten Pfad verwenden, um an einer beliebigen Stelle im System die Ausgabe abzulegen. - + Plain Text Einfacher Text - + When checked, a review window will open upon process completion. Wenn diese Option aktiviert ist, wird nach Abschluss des Prozesses ein Überprüfungsfenster geöffnet. - + Output File: Ausgabedatei: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Geben Sie entweder einen absoluten oder relativen Pfad mit einem Dateinamen an. Die Ausgabe aller Bilder wird in dieselbe Datei geschrieben.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. erkannter_text.txt - + Cancel Action Aktion Abbrechen - + Close Schließen - + Generating: This is present progressive, as in "[currently] Generating:" Erzeugt: - + Current Step: Aktueller Schritt: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Bilder Verarbeitet: - + Set Default Standard festlegen - + Settings Einstellungen - + Theme As in color theme Farbschema - + Language Sprache - + Help Hilfe - + Add Files... Dateien hinzufügen... - + Remove All Files Alle Dateien entfernen - + Remove File Datei entfernen - + Help Translate Panel Cleaner Helfe, Panel Cleaner zu übersetzen - + Report an Issue... Melde ein Problem... - + Simulate Exception Error simulieren - + Add File Browser Extension... Dateibrowser-Erweiterung hinzufügen... - + Show Supported OCR Languages Unterstützte OCR-Sprachen anzeigen - - + + Reset Window Layout Preferences Fensterlayout-Einstellungen zurücksetzen - + Show Out Of Memory Warnings Bei geringem Arbeitsspeicher warnen - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Nachlaufaktionen - + Configure Post-Run Actions... Nachlaufaktionen konfigurieren... - + New Neu - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>Wenn diese Option aktiviert ist, werden die Ausgaben auf der Festplatte gespeichert. Andernfalls können Sie sie nur in der Bilddetailansicht oder in der Ausgabeüberprüfung (falls aktiviert) ansehen. </p><p>Wenn das Profil nach der Reinigung unverändert bleibt, ohne dass diese Option aktiviert ist, können Sie sie schnell exportieren, indem Sie den Reininger erneut mit aktivierter Option ausführen.</p></body></html> - + CSV Spreadsheet CSV-Tabelle - + Delete Löschen - + Import... Importieren... - + Online Documentation Online-Dokumentation - + About Panel Cleaner Über Panel Cleaner - + View License Lizenz anzeigen - + Save as... Speichern unter... - + System System - + Dark Dunkel - + Light Hell - + Add Folder... Ordner hinzufügen... - + Delete Machine Learning Models Modelle für maschinelles Lernen löschen - + Download Machine Learning Models Modelle für maschinelles Lernen herunterladen - + Donate Spenden @@ -2139,19 +2144,19 @@ Lesen Sie in der Online-Dokumentation nach, wie Sie die Unterstützung für weit OCR-Modell: - + Already downloaded Bereits heruntergeladen - + Text Detector model (CUDA) Texterkennungsmodell (CUDA) - + Text Detector model (CPU) Texterkennungsmodell (CPU) @@ -2170,7 +2175,7 @@ Lesen Sie in der Online-Dokumentation nach, wie Sie die Unterstützung für weit OCR-Download-Fehler - + ETA estimated time of completion Restzeit @@ -2748,7 +2753,7 @@ Lesen Sie in der Online-Dokumentation nach, wie Sie die Unterstützung für weit Der Name "{shutdown_command}" ist reserviert. - + The Flatpak sandbox will likely prevent actions from working. Die Flatpak-Sandbox wird wahrscheinlich verhindern, dass die Aktionen funktionieren. @@ -2829,73 +2834,73 @@ Lesen Sie in der Online-Dokumentation nach, wie Sie die Unterstützung für weit Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input Das Originalbild mit angewandter Skalierung (falls erforderlich). - + Input Step name in the image details view Eingabe - + Input Output name in the image details view Eingabe - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask Die von der KI generierte Grobmaske. - + Text Detection Step name in the image details view Text-Erkennung - + AI Mask Output name in the image details view KI-Maske - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes Die von der KI erzeugten ungefilterten Feld-Daten. - + Raw Boxes Output name in the image details view Rohe Felder - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes Die Umrisse der Textfelder, die die KI gefunden hat. - + Preprocessor Step name in the image details view Präprozessor - + Initial Boxes Output name in the image details view Initiale Felder - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2903,181 +2908,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Grün: initiale Felder. Rot: erweiterte Felder. Lila: Zusammengefügte (finale) Felder. Blau: Referenzfelder für die Entrauschung. - + Final Boxes Output name in the image details view Finale Felder - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask Die Maske der zusammengefügten Felder. - + Masker Step name in the image details view Maskierer - + Box Mask Output name in the image details view Felder-Maske - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask Die grobe Texterkennungsmaske mit allem, was außerhalb der Felder-Maske liegt, weggeschnitten. - + Cut Mask Output name in the image details view Schnittmaske - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers Die verschiedenen Wachstumsstufen rund um die Schnittmaske in verschiedenen Farben dargestellt. - + Mask Layers Output name in the image details view Maskenstufen - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality Die Standardabweichung (σ) und die Umrissdicke (in Pixeln) jeder ausgewählten besten Maske, falls vorhanden. Ein geringeres σ ist besser, von perfekt (lila) bis fehlgeschlagen (rot). - + Fitment Quality Output name in the image details view Anpassungsqualität - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay Das Eingabebild mit der finalen Maske in Farbe überlagert. - + Mask Overlay Output name in the image details view Maskenüberlagerung - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask Die Sammlung von Masken für jede Blase, die am besten passen. - + Final Mask Output name in the image details view Finale Maske - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text Die aus dem Eingabebild isolierte Textebene. - + Isolated Text Output name in the image details view Isolierter Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output Das Eingabebild mit der finalen Maske überlagert. - + Masked Output Output name in the image details view Maskierte Ausgabe - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask Die Masken, die entrauscht werden müssen. Sie werden beim Exportieren über die finale Maske gelegt. - + Denoiser Step name in the image details view Entrauscher - + Denoise Mask Output name in the image details view Entrauschungsmaske - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output Das Eingabebild mit der entrauschten Maske überlagert. - + Denoised Output Output name in the image details view Entrauschte Ausgabe - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask Die übermalten Abschnitte um den Text herum, die, wenn überhaupt, nur schlecht gereinigt wurden. - + Inpainter Step name in the image details view Übermaler - + Inpainted Mask Output name in the image details view Übermalte Maske - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output Das Eingabebild: gereinigt, entrauscht (falls aktiviert) und übermalt. - + Inpainted Output Output name in the image details view Übermalte Ausgabe - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3086,1141 +3091,1141 @@ Grün: initiale Felder. Rot: erweiterte Felder. Lila: Zusammengefügte (finale) Profile - + General Profile section title Allgemein - + About this profile: Profile comment in General Über dieses Profil: - + Notes Profile entry in General Notizen - + Preferred file type to save the cleaned image as. Profile comment in General Bevorzugter Dateityp, unter dem das bereinigte Bild gespeichert werden soll. - + Preferred File Type Profile entry in General Bevorzugter Dateityp - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Bevorzugter Dateityp, unter dem die Maske gespeichert werden soll. Es werden nur Bildformate unterstützt, die Transparenz zulassen. - + Preferred Mask File Type Profile entry in General Bevorzugter Maskendateityp - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General Im Folgenden sind die unteren und oberen Zielwerte für die Höhe des Eingabebildes aufgeführt. Es wird immer nur so weit verkleinert, dass es in den Bereich passt, wobei ganzzahlige Faktoren bevorzugt werden, um die Auswirkungen auf die Bildqualität zu minimieren. Bilder, die kleiner als der obere Zielwerte sind, bleiben unverändert. Sie können diese Funktion deaktivieren, indem Sie einen oder beide Werte kleiner als oder gleich 0 setzen. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General Dies ist nützlich, um die Verarbeitung von großen Bildern erheblich zu beschleunigen. Da andere Optionen, die sich auf die Pixeldimensionen stützen, von der Größe abhängen, hilft dies außerdem, die Ergebnisse für verschiedene Bildgrößen zu normalisieren. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General Das Bild wird verkleinert, verarbeitet und dann wird nur die Maske wieder hochskaliert. Das bedeutet, dass für die bereinigte Ausgabe weiterhin das ursprüngliche, nicht skalierte Bild verwendet wird, um Qualitätsverluste zu verhindern. Zur Bestimmung des Skalierungsfaktors wird nur die Höhe des Bildes betrachtet, wobei das Seitenverhältnis erhalten bleibt und die individuelle Breite eines Bildes ignoriert wird, so dass der Faktor konsistent bleibt, wenn es sich bei einem Bild um eine Doppelseite handelt. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General Z.B. für ein unteres Ziel von 1000 und ein oberes Ziel von 2000 wird ein Bild mit der Größe 5000x7000 (B, H) Pixel um den Faktor 4 verkleinert, sodass es bei der Verarbeitung die Größe 1250x1750 Pixel hat. - + Input Height Lower Target Profile entry in General Eingabehöhe unteres Ziel - + Input Height Upper Target Profile entry in General Eingabehöhe oberes Ziel - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Lange Streifen in einzelne Abschnitte aufteilen. Wenn diese Option aktiviert ist, wird der gesamte Streifen nicht auf die gewünschte Höhe gequetscht, sondern in einzelne Abschnitte aufgeteilt, die jeweils in die gewünschte Höhe plus/minus das Doppelte der Teilungstoleranz passen (wenn das Segment in der Mitte des Streifens liegt). - + Split Long Strips Profile entry in General Lange Streifen Teilen - + Preferred height to split long strips at. Profile comment in General Bevorzugte Länge zum Spalten langer Streifen. - + Preferred Split Height Profile entry in General Bevorzugte Spalt-Länge - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Toleranzspanne für die Teilung langer Streifen. Dies ist die maximale Differenz zwischen der gewünschten Teilungslänge und der tatsächlichen Länge des Streifens. Ein Algorithmus bestimmt den besten Teilungspunkt innerhalb dieser Spanne, um eine Teilung in der Mitte von etwas Wichtigem zu vermeiden. - + Split Tolerance Margin Profile entry in General Spalt-Toleranzspanne - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Seitenverhältnis, das für die Aufteilung langer Streifen verwendet wird. Dies ist das Verhältnis der Breite zur Höhe des Bildes. Wenn das Seitenverhältnis des Bildes kleiner als dieser Wert ist, wird es als langer Streifen betrachtet.<br> Beispiel: 0,3 bedeutet, dass die Breite des Bildes das 0,3-fache der Höhe beträgt. - + Long Strip Aspect Ratio Profile entry in General Kritisches Seitenverhältnis - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Lange Streifen wieder zu einem einzigen Bild zusammenfügen. Wenn diese Option aktiviert ist, werden die einzelnen Seiten, die aus einem langen Streifen erstellt wurden, beim Export wieder zu einem einzigen Bild zusammengefügt. - + Merge After Split Profile entry in General Streifen Zusammenfügen - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximale Anzahl von Kernen, die für den Export von Bildern verwendet werden sollen. Sie können diesen Wert unbestimmt lassen, um alle verfügbaren Kerne zu verwenden. Verringern Sie diesen Wert, wenn Sie auf Arbeitsspeicherprobleme stoßen, die zu zufälligen abstürzen führen. - + Max Threads Export Profile entry in General Max Kerne Export - + Text Detector Profile section title Text-Erkennung - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Pfad zum Texterkennungsmodell. Leer lassen, um das integrierte Modell zu verwenden. Es wird nur empfohlen, dies zu überschreiben, wenn die automatisch heruntergeladene Version älter ist als die neueste Version. Der Pfad muss direkt auf die Datei comictextdetector.pt (CUDA) oder comictextdetector.pt.onnx (CPU) verweisen. Sie können ältere Versionen des Modells <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">hier herunterladen.</a> - + Model Path Profile entry in Text Detector Modell Pfad - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Anzahl der Modelle, die parallel ausgeführt werden sollen. Dies ist nützlich, wenn Sie über genügend RAM (oder VRAM bei CUDA) verfügen, um mehrere Modelle gleichzeitig auszuführen. Dies erhöht natürlich die Geschwindigkeit des Prozesses, kann aber auch Ihren Computer zum Absturz bringen, wenn Sie Ihre Hardware überschätzen. Ich empfehle, 1 Modell pro 2 GB verfügbaren Speicher oder 1 Modell pro 50 Bilder zu verwenden. Hinweis: Dies wird ignoriert, wenn weniger als 50 Dateien verarbeitet werden, da sich der Overhead des Starts mehrerer Modelle darunter nicht lohnt. Warnung: Dies kann zur Instabilität des Programms führen, die Verwendung erfolgt auf eigene Gefahr. <br>Berichten Sie NICHT über Probleme mit dieser Einstellung, da sie vollständig hardwareabhängig ist! - + Concurrent Models Profile entry in Text Detector Parallele Modelle - + Preprocessor Profile section title Präprozessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Die Größe der Felder wird in der Gesamtzahl der Pixel angegeben. Ein Feld von 200x200 Pixeln hat also eine Größe von 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Mindestgröße eines Feldes, damit es nicht verworfen wird. - + Box Min Size Profile entry in Preprocessor Min. Feldgröße - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Mindestgröße eines Feldes mit "unbekannter" Sprache, um es nicht zu verwerfen. Diese Sprache wird in der Regel für Logos und andere Textelemente verwendet, wo die KI nicht weiß, was sie mit ihnen anfangen soll. - + Suspicious Box Min Size Profile entry in Preprocessor Verdächtige Min. Feldgröße - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor Die Mindestüberlappung zwischen zwei Feldern, um sie zusammenzuführen. Dieser Prozentsatz (0-100) gibt an, wie viel des kleineren Feldes innerhalb des größeren Feldes liegen muss, um zusammengeführt zu werden. Bei einem höheren Wert ist eine größere Überlappung erforderlich, um die Felder zusammenzuführen. - + Box Overlap Threshold Profile entry in Preprocessor Min. Feldüberlappung - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Ob OCR verwendet werden soll, um Felder zu erkennen, die nicht bereinigt werden sollten, z. B. Felder, die nur Zahlen oder Symbole enthalten. - + OCR Enabled Profile entry in Preprocessor OCR Aktiviert - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Ob Tesseract zur Durchführung von OCR-Aufgaben verwendet werden soll.<br> Wenn hier ein Häckchen gesetzt ist, kann Tesseract OCR für die Textextraktion verwendet werden, sofern verfügbar.<br> Wenn dieses Häckchen nicht gesetzt ist, wird immer das integrierte OCR-Modell (manga-ocr) verwendet, das sich am besten für vertikalen japanischen Text eignet. - + OCR Use Tesseract Profile entry in Preprocessor OCR Tesseract verwenden - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor Die Sprache, die für OCR-Aufgaben verwendet werden soll. Die Texterkennung kann Japanisch und Englisch erkennen, für alle anderen Sprachen wählen Sie hier explizit eine Sprache aus. Bei der Erkennung pro Feld wird beibehalten, was der Texterkenner erkannt hat, während bei der Erkennung pro Seite die häufigste Sprache der gesamten Seite zugewiesen wird.<br> Hinweis: Nur Japanisch wird standardmäßig unterstützt, für alle anderen Sprachen müssen Tesseract aktiviert und die zugehörigen Sprachpakete installiert sein. - + OCR Language Profile entry in Preprocessor OCR-Sprache - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Gibt an, welche Engine für die OCR verwendet werden soll.<br> - auto: Wählt automatisch die OCR-Engine auf der Grundlage der erkannten Sprache jedes Textblocks im Bild aus. Verwendet Manga Ocr für japanischen Text, Tesseract für englischen oder unbekannten Text.<br> - mangaocr: Zwingt Panel Cleaner, das eingebaute Manga-Ocr-Modell für alle Texterkennungsaufgaben zu verwenden. Am besten geeignet für vertikalen japanischen Text.<br> - tesseract: Zwingt Panel Cleaner zur Verwendung von Tesseract OCR für alle Texterkennungsaufgaben. Dies ist eine vielseitige Option, die Englisch und mehrere andere Sprachen unterstützt. - + OCR Engine Profile entry in Preprocessor OCR-Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Legt die Lesereihenfolge für die Verarbeitung und Sortierung von Textfeldern auf der gesamten Seite fest, nicht für einzelne Textblöcke. Diese globale Einstellung beeinflusst, wie Textfelder geordnet und für die weitere Verarbeitung dargestellt werden.<br> - auto: Erkennt die Lesereihenfolge anhand der erkannten Sprache der einzelnen Textblöcke auf der Seite.<br> - manga: Reihenfolge von rechts nach links, von oben nach unten. Geeignet für japanische Manga.<br> - comic: Von links nach rechts, von oben nach unten. Geeignet für westliche Comics und Texte.<br> Wählen Sie nach dem vorherrschenden Layout Ihres Inhalts. - + Reading Order Profile entry in Preprocessor Reihenfolge - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximale Größe eines Feldes, für das OCR durchgeführt werden soll. Die Unsinnsfelder sind in der Regel klein und die OCR ist langsam, daher sollten Sie dies als Grenzwert verwenden. - + OCR Max Size Profile entry in Preprocessor OCR Max. Größe - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex-Muster zum Abgleich mit OCR-Ergebnissen. Alles, was mit diesem Muster übereinstimmt, wird verworfen.<br> Beachten Sie, dass das MangaOCR-Modell Sonderzeichen in voller Breite zurückgibt. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR-Blacklist-Muster - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor Das MangaOCR-Modell kann nur japanischen Text verarbeiten. Wenn also die Option Strenge Spracherkennung aktiviert ist, werden Felder verworfen, bei denen die Texterkennung nicht sicher ist, dass sie japanisch enthalten'. Manchmal verunsichern Zahlen oder andere Symbole die KI, was dazu führt, dass die erkannte Sprache unbekannt ist. Wenn Streng deaktiviert ist, werden diese nicht verworfen. Alles, was mit Sicherheit als eine andere Sprache erkannt wird, wird unabhängig von dieser Einstellung verworfen.<br> Hinweis: Diese Einstellung ist nur relevant, wenn ocr_language auf die Erkennung pro Feld oder Seite eingestellt ist. - + OCR Strict Language Profile entry in Preprocessor OCR Strenge Spracherkennung - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Füllung, die zu jeder Seite eines Feldes hinzugefügt wird. Dies wird zu den initialen Feldern hinzugefügt, die durch die Texterkennungs-KI erstellt wurden. Diese Rahmen werden in Grün dargestellt. - + Box Padding Initial Profile entry in Preprocessor Feldauffüllung Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Füllung für die rechte Seite eines Feldes. Diese Erweiterung hilft dabei, Rubytext zu verdecken, der rechts von vertikalem Text übersteht. - + Box Right Padding Initial Profile entry in Preprocessor Feldauffüllung Rechts Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Füllung, die zu jeder Seite eines Feldes hinzugefügt wird. Dies wird zu einem erweiterten Satz von Feldern hinzugefügt, die verwendet werden, um Falschmeldungen durch die KI-Maske des Texterkenners auszuschließen. Diese Kästchen werden in lila dargestellt. - + Box Padding Extended Profile entry in Preprocessor Feldauffüllung Erweitert - + Box Right Padding Extended Profile entry in Preprocessor Feldauffüllung Rechts Erweitert - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Füllung, die zu jeder Seite eines Feldes hinzugefügt wird. Dies wird zu den Referenzfeldern hinzugefügt, die bei der Analyse der zu verwendenden Maske zum Abtasten des Originalbildes verwendet werden. Diese Rahmen werden in Blau dargestellt. - + Box Reference Padding Profile entry in Preprocessor Feldauffüllung Referenz - + Masker Profile section title Maskierer - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximale Anzahl von Kernen, die für die Maskenerzeugung verwendet werden sollen. Sie können diesen Wert unbestimmt lassen, um alle verfügbaren Kerne zu verwenden. Verringern Sie diesen Wert, wenn Sie auf Arbeitsspeicherprobleme stoßen, die zu zufälligen abstürzen führen. - + Max Threads Profile entry in Masker Max Kerne - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Anzahl der Pixel, um die die Maske bei jedem Schritt vergrößert wird. Dadurch wird der Umriss der Maske vergrößert. Kleinere Werte sind also genauer, aber langsamer. - + Mask Growth Step Pixels Profile entry in Masker Maskenwachstum pro Schritt - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Anzahl der Schritte, um die die Maske wachsen soll. Bei einer höheren Zahl werden mehr und immer größere Masken erstellt, die letztlich durch die Größe des Referenzfeldes begrenzt sind. - + Mask Growth Steps Profile entry in Masker Maskenwachstumsschritte - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Mindestdicke einer Maske. Dies entspricht dem ersten Wachstumsschritt der Maske, die übrigen folgen dem Wachstumsschritt der Maske in Pixeln. Auf diese Weise können Sie eine kleine Schrittgröße für die Genauigkeit haben, aber immer noch verhindern, dass sehr dünne Masken für Text verwendet werden, der nur einen Umriss hat, aber sonst vor einem Hintergrund ist, d.h. dort wäre Inpainting besser gewesen. - + Min Mask Thickness Profile entry in Masker Min Maskendicke - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter Die minimale Standardabweichung der Farben um den Rand einer bestimmten Maske, um den Bereich um die Maske herum zu übermalen. Wenn dieser Wert mit der maximalen Abweichung für Masken übereinstimmt, werden nur fehlgeschlagene Felder übermalt, sodass die folgenden beiden Einstellungen irrelevant sind. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Kombinieren Sie die Ausgaben in einer einzigen Projektdatei als Ebenen. Derzeit unterstützte Formate: Photoshop PSD.<br> - Keine: Jedes Bild und jede Maske wird als einfache Bilddatei gespeichert.<br> - PSD Pro Bild: Bilder und Masken werden zusammen in einer PSD-Datei pro Eingabebild gespeichert.<br> - PSD Sammeldatei: Alle Bilder und Masken werden zusammen in einer einzigen PSD-Datei gespeichert, gruppiert nach Eingabebild. - + Layered Export Profile entry in General Ebenenexport - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Ob farbige Masken zugelassen werden sollen. Wenn diese Option auf ausgewählt ist, können Masken in jeder beliebigen Farbe erstellt werden, nicht nur in weiß, schwarz oder grau. - + Allow Colored Masks Profile entry in Masker Farbige Masken zulassen - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximaler Schwellenwert, ab dem ein Pixel nicht als weiß gilt. Die Medianfarbe entlang des Randes einer Maske ist möglicherweise nicht reinweiß. Um leichte gebrochene Weißtöne auf einem reinweißen Hintergrund zu vermeiden, wird alles, was heller als dieser Schwellenwert ist, auf reinweiß aufgerundet. Wertebereich: Schwarz (0) bis Reinweiß (255). - + Off White Max Threshold Profile entry in Masker Max. Schwellenwert für gebrochenes Weiß - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker Die Standardabweichung einer Maske stellt die Farbabweichung entlang des Randes der Maske dar. Dazu werden nur die einzelnen Pixel entlang des äußeren Randes einer Maske aus dem Originalbild abgetastet. Eine geringe Abweichung bedeutet, dass die Maske auf einem gleichmäßigen Hintergrund liegt, d. h. sie überschneidet keinen Text oder andere Objekte. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimale Verbesserung der Standardabweichung der Maske, um sie weiter zu vergrößern. Wird hier ein höherer Wert eingestellt, so ist eine höhere Verbesserung erforderlich, um eine größere Maske zu berücksichtigen, um kleineren Masken den Vorzug zu geben. - + Mask Improvement Threshold Profile entry in Masker Maskenverbesserungsschwellwert - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Auswahl, ob der schnelle Maskenauswahlalgorithmus verwendet werden soll. Wenn diese Option aktiviert ist, wählt der Algorithmus für die Maskenauswahl die erste perfekte Maske aus, wenn eine solche früh gefunden wird. Dies ist schneller, aber es kann sein, dass nicht die beste Maske gefunden wird, wenn eine etwas größere besser gewesen wäre. - + Mask Selection Fast Profile entry in Masker Schnelle Maskenauswahl - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximale Anzahl von Kernen, die für die Entrauschung verwendet werden sollen. Sie können diesen Wert unbestimmt lassen, um alle verfügbaren Kerne zu verwenden. Verringern Sie diesen Wert, wenn Sie auf Arbeitsspeicherprobleme stoßen, die zu zufälligen abstürzen führen. - + Max Threads Profile entry in Denoiser Max Kerne - + Inpainter Profile section title Übermaler - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTELL: Wenn Sie bessere Standardeinstellungen finden, öffnen Sie bitte ein Issue auf Github, um die Verbesserungen mit allen zu teilen. Beachten Sie, dass Übermalung nicht standardmäßig aktiviert ist. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Beim Übermalen wird maschinelles Lernen eingesetzt, um den Inhalt eines Bildes auf der Grundlage seiner Umgebung zu ersetzen. Für Masken, die nicht gut (oder gar nicht) bereinigt werden konnten, kann Übermalung verwendet werden. Zur Unterscheidung von der Entrauschung ist das Übermalen für deutlich schlechtere Masken gedacht, die eine enge Passung aufweisen. Masken, die entrauscht wurden, werden nicht übermalt. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Da dieser Schritt in einigen Fällen zu schlechten Ergebnissen führen kann, kann er hier deaktiviert werden. - + Inpainting Enabled Profile entry in Inpainter Übermalen Aktiviert - + Inpainting Min Std Dev Profile entry in Inpainter Übermalung Min. Abweichung - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter Der maximale Radius einer Maske für das Übermalen. Masken, die dicker sind als dieser Wert, werden unverändert belassen, denn wenn der Rand so groß ist, ist die Maske wahrscheinlich schon gut genug. - + Inpainting Max Mask Radius Profile entry in Inpainter Übermalung Max. Maskenradius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter Der Mindestradius um eine zu übermalende Maske. Dieser Wert wird zur optimalen Maskengröße addiert, um sicherzustellen, dass die Übermalung die gesamte Maske abdeckt. - + Min Inpainting Radius Profile entry in Inpainter Min. Übermalungsradius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter Bei Masken, die sich als wesentlich schwieriger zu bereinigen erweisen, d. h. eine hohe Standardabweichung aufweisen, wird der Radius der Übermalung erhöht, um einen größeren Teil der Maske abzudecken. Dieser zusätzliche Spielraum wird zum minimalen Übermalungsradius addiert und wie folgt berechnet: Übermalungsradius-Faktor mal Standardabweichung der Maske - + Inpainting Radius Multiplier Profile entry in Inpainter Übermalungsradius-Faktor - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter Der maximale Radius um eine zu übermalende Maske. Dies begrenzt die Größe, bis zu der die Übermalung wachsen kann, um zu verhindern, dass sie zu viel des Bildes bedeckt, wenn ein großer Übermalungsradius-Faktor verwendet wird. - + Max Inpainting Radius Profile entry in Inpainter Max. Übermalungsradius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter Nach dem Übermalen wird ein begrenzter Bereich aus dem Originalbild ausgeschnitten, um zu verhindern, dass sich das Übermalen auf den Rest des Bildes auswirkt. Dadurch wird sichergestellt, dass das Originalbild so weit wie möglich erhalten bleibt. Dieser Radius wird um den endgültigen Übermalungsradius herum hinzugefügt, da das Übermalungsmodell einige Pixel außerhalb des ihm zugewiesenen Bereichs verändert. - + Inpainting Isolation Radius Profile entry in Inpainter Übermalung-Isolationsradius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Verblasst die Ränder der übermalten Abdeckmaske um diese Anzahl von Pixeln aus, um die übermalten Teile sanft in den Rest des Bildes zu integrieren.<br> Wenn Sie nach dem Übermalen schwache Umrisse sehen, erhöhen Sie den Mindestwert für das Übermalen und setzen Sie diesen Wert auf die Hälfte dieses Wertes. - + Inpainting Fade Radius Profile entry in Inpainter Übermalung Verblassungsradius - + Detect per box Language option for OCR Erkennen pro Feld - + Detect per page Language option for OCR Erkennen pro Seite - + Japanese Language option for OCR Japanisch - + English Language option for OCR Englisch - + Korean Language option for OCR Koreanisch - + Korean (vertical) Language option for OCR Koreanisch (vertikal) - + Chinese - Simplified Language option for OCR Chinesisch - Vereinfacht - + Chinese - Traditional Language option for OCR Chinesisch - Traditionell - + Albanian Language option for OCR Albanisch - + Arabic Language option for OCR Arabisch - + Azerbaijani Language option for OCR Aserbaidschanisch - + Azerbaijani - Cyrilic Language option for OCR Aserbaidschanisch - Kyrillisch - + Bengali Language option for OCR Bengalisch - + Bulgarian Language option for OCR Bulgarisch - + Burmese Language option for OCR Birmanisch - + Catalan; Valencian Language option for OCR Katalanisch; Valencianisch - + Croatian Language option for OCR Kroatisch - + Czech Language option for OCR Tschechisch - + Danish Language option for OCR Dänisch - + Dutch; Flemish Language option for OCR Niederländisch; Flämisch - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estnisch - + Finnish Filipino Language option for OCR Finnisch - + French Language option for OCR Französisch - + Georgian Language option for OCR Georgisch - + German Language option for OCR Deutsch - + Greek Language option for OCR Griechisch - + Hebrew Language option for OCR Hebräisch - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Ungarisch - + Indonesian Language option for OCR Indonesisch - + Italian Language option for OCR Italienisch - + Kazakh Language option for OCR Kasachisch - + Latin Language option for OCR Latein - + Lithuanian Language option for OCR Litauisch - + Luxembourgish Language option for OCR Luxemburgisch - + Malay Language option for OCR Malaiisch - + Mongolian Language option for OCR Mongolisch - + Nepali Language option for OCR Nepalesisch - + Norwegian Language option for OCR Norwegisch - + Persian Language option for OCR Persisch - + Polish Language option for OCR Polnisch - + Portuguese Language option for OCR Portugiesisch - + Romanian; Moldavian Language option for OCR Rumänisch; Moldawisch - + Russian Language option for OCR Russisch - + Serbian Language option for OCR Serbisch - + Serbian - Latin Language option for OCR Serbisch - Latein - + Slovak Language option for OCR Slowakisch - + Slovenian Language option for OCR Slowenisch - + Spanish; Castilian Language option for OCR Spanisch; Kastilisch - + Swedish Language option for OCR Schwedisch - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamilisch - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thailändisch - + Turkish Language option for OCR Türkisch - + Ukrainian Language option for OCR Ukrainisch - + Vietnamese Language option for OCR Vietnamesisch - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker Die maximale Standardabweichung einer Maske, die berücksichtigt werden soll. Ein hoher Wert bedeutet eine höhere Toleranz für eine Maske, die Text oder andere Objekte überschneidet, was keine gute Maske ist, da dieses Feld ohnehin manuelle Korrekturen benötigen wird. Wenn Sie diesen Wert auf 0 setzen, werden nur perfekte Masken zugelassen, was für Bilder mit sehr hoher Auflösung empfohlen wird. - + Mask Max Standard Deviation Profile entry in Masker Max. Standardabweichung einer Maske - + Color to use for the debug mask. Profile comment in Masker Farbe, die für die Debug-Maske verwendet werden soll. - + Debug Mask Color Profile entry in Masker Farbe der Debug-Maske - + Denoiser Profile section title Entrauscher - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser Wenn ein Stück Text schwer zu maskieren ist, musste bei der Bereinigung wahrscheinlich eine kleine Maske gewählt werden, die ggf. eine Menge JPEG-Artefakte um den Text herum zurücklässt. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser Das bedeutet, dass die Farbe der Pixel um den Rand einer Maske herum nicht einheitlich ist, was als Standardabweichung quantifiziert wird. Der Entrauscher kann auf Masken ab einer minimalen Standardabweichung abzielen und den Bereich unmittelbar um die Maske herum entrauschen. Dadurch bleiben die Details im restlichen Bild erhalten, aber die Artefakte im Bereich des ehemaligen Textes werden entfernt. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Da es sich hierbei um einen optionalen Schritt handelt, der bei hochaufgelösten Bildern, die nicht unter JPEG-Artefakten leiden, sogar überflüssig ist, kann er hier deaktiviert werden. - + Denoising Enabled Profile entry in Denoiser Entrauschen Aktiviert - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser Die minimale Standardabweichung der Farben um den Rand einer gegebenen Maske, um eine Entrauschung in dem Bereich um die Maske durchzuführen. - + Noise Min Standard Deviation Profile entry in Denoiser Rauschen Min. Standardabweichung - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Hinweis: Wenn die Übermalung aktiviert ist, gilt die minimale Standardabweichung für die Übermalung als Höchstwert, nach dem diese Maske an den Übermaler weitergegeben wird. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser Die Dicke eines Umrisses, der um eine Maske herum entrauscht werden soll. - + Noise Outline Size Profile entry in Denoiser Rauschumrissgröße - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Blenden Sie die Ränder der entrauschten Abdeckmaske um diesen Wert ab, um die entrauschten Teile sanft in den Rest des Bildes zu integrieren. - + Noise Fade Radius Profile entry in Denoiser Rauschblendenradius - + Set to True to support colored images. Profile comment in Denoiser Bilder mit Farbe entrauschen. - + Colored Images Profile entry in Denoiser Farbige Bilder - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser Wie stark das Bild entrauscht werden soll. Höhere Werte bedeuten eine stärkere Rauschunterdrückung. - + Filter Strength Profile entry in Denoiser Filterstärke - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser Wie stark die Farbkanäle entrauscht werden sollen, im Gegensatz zur Helligkeit. Höhere Werte bedeuten mehr Rauschunterdrückung. - + Color Filter Strength Profile entry in Denoiser Farbfilterstärke - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Größe des Kontextfensters in Pixeln, das für die Berechnung der Gewichte verwendet wird. Sollte ungerade sein. - + Template Window Size Profile entry in Denoiser Kontextfenstergröße - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Größe des Fensters in Pixeln, das zur Berechnung des gewichteten Durchschnitts für ein bestimmtes Pixel verwendet wird. Sollte ungerade sein. - + Search Window Size Profile entry in Denoiser Suchfenstergröße @@ -4247,7 +4252,7 @@ Grün: initiale Felder. Rot: erweiterte Felder. Lila: Zusammengefügte (finale) px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. Alle Kerne @@ -4664,14 +4669,14 @@ Grün: initiale Felder. Rot: erweiterte Felder. Lila: Zusammengefügte (finale) ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>Der Pfad für das Text-Detektor-Modell {value} existiert nicht und wird auf die Standardeinstellungen zurückgesetzt. Sie können das Modell manuell <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">hier</a> herunterladen oder weiterhin das Standardmodell verwenden.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. Das Regex-Muster "{value}" für {key} ist ungültig und wird auf den Standardwert zurückgesetzt. From 4b4a5aada68bfad484e6cfe2360cc1a4a5c151be Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:15 +0100 Subject: [PATCH 10/18] New translations panelcleaner_source.ts (Italian) --- translations/PanelCleaner_it_IT.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_it_IT.ts b/translations/PanelCleaner_it_IT.ts index 2e4183b..0742fac 100644 --- a/translations/PanelCleaner_it_IT.ts +++ b/translations/PanelCleaner_it_IT.ts @@ -4,7 +4,7 @@ - + default reserved profile name predefinito @@ -20,54 +20,54 @@ Profilo {profile_name} creato. - + Profile name cannot be empty. Il nome del profilo non può essere vuoto. - + Profile name already in use. Nome del profilo già in uso. - + Profile name is reserved. Il nome del profilo non è disponibile. - - - - + + + + Unknown Sconosciuto - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Scala di grigi - + 1-bit Color mode 1-bit - + Palette Color mode Palette @@ -78,45 +78,45 @@ Errore durante il download del file dall'url: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Errore nello scaricare il file dall'url: {url} Failed to save the file to {save_path} - + Error downloading file from url: {url} The file content is different from expected. Errore nello scaricare il file dall'url: {url} Il contenuto del file è diverso da quello previsto. + - - - + + + - - - + + N/A N/A - + OCR Analytics Analisi OCR - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Numero di riquadri: {num_boxes} | Numero di riquadri piccoli: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Numero di riquadri rimossi: {num_removed} ({removed_box_ratio} totale, {removed_among_small_ratio} di riquadri piccoli) @@ -155,66 +155,66 @@ Fumetti puliti: - + Small boxes Riquadri piccoli - + Removed boxes Fumetti puliti - - + + Failed Operazione non riuscita - + Perfect Perfetto - + Total Totale - + Mask Fitment Analytics Analisi dell'applicazione della maschera - + Total Totale - + Total boxes Box in totale - + Masks succeeded Maschere riuscite - + Masks failed Maschere fallite - + Perfect masks Maschere perfette - + Average border deviation Deviazione media del confine @@ -233,7 +233,7 @@ Utilizzo della maschera per spessore (in pixel): Maschera ({thickness}px) - + Box mask Maschera box @@ -254,27 +254,27 @@ Pagine con errori / totale: - + Denoising Analytics Analisi del denoising - + Total masks Totale maschere - + Masks denoised Maschere col denoising - + Minimum deviation to denoise Deviazione minima per il denoising - + Maximum allowed deviation Deviazione massima consentita @@ -293,27 +293,27 @@ Pagine con errori / totale: Denoised - + Inpainting Analytics Analitica dell'Inpainting - + Inpainting performed Inpainting eseguito - + Average thickness Spessore medio - + Minimum thickness Spessore minimo - + Maximum thickness Spessore massimo @@ -372,7 +372,7 @@ Pagine con errori / totale: Il formato del file non è stato riconosciuto. - + The CSV file must start with a header row, followed by data rows. Il file CSV deve iniziare con una riga di intestazione, seguita da righe di dati. @@ -402,7 +402,7 @@ Pagine con errori / totale: Più immagini corrispondono al percorso del file: {path} - + Error, no version found. Error, no version found. @@ -412,7 +412,7 @@ Pagine con errori / totale: CUDA Error - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pagine con errori / totale: Il modello di Inpainting non è stato trovato. - + filename nome del file - + startx startx - + starty starty - + endx endx - + endy endy - + text testo - + Saved detected text to {output_file} Salvataggio del testo rilevato in {output_file} - + Failed to write detected text to {output_file} Impossibile scrivere il testo rilevato su {output_file} - + Save Failed Salvataggio fallito - + Failed to write detected text to file. Impossibile scrivere il testo rilevato sul file. @@ -538,7 +538,7 @@ Pagine con errori / totale: - + Page layered export @@ -625,7 +625,7 @@ Uno strumento basato sull'intelligenza artificiale per pulire i pannelli dei man ColorButton - + Select Color Seleziona il colore @@ -671,7 +671,7 @@ Uno strumento basato sull'intelligenza artificiale per pulire i pannelli dei man Sistema non supportato - + This system is not supported. Questo sistema non è supportato. @@ -686,12 +686,12 @@ Uno strumento basato sull'intelligenza artificiale per pulire i pannelli dei man Disinstallazione per Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. L'estensione verrà installata all'indirizzo {path}. Questo'non richiede i permessi di root. - + The extension file will be deleted. This doesn't require root permissions. Il file di estensione verrà eliminato. Questa operazione non'richiede i permessi di root. @@ -706,17 +706,17 @@ Uno strumento basato sull'intelligenza artificiale per pulire i pannelli dei man Disinstallazione per Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. L'estensione verrà installata nel registro di Windows e richiede i permessi di amministratore. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. L'estensione verrà disinstallata dal registro di Windows. Questa operazione richiede i permessi di amministratore. - + Dolphin extension installed successfully. L'estensione Dolphin è stata installata correttamente. @@ -726,7 +726,7 @@ Uno strumento basato sull'intelligenza artificiale per pulire i pannelli dei man Installazione riuscita - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. L'estensione Dolphin è stata installata correttamente. @@ -806,12 +806,12 @@ Per impostazione predefinita, utilizza il profilo incorporato. Tuttavia, è poss FileTable - + Original size in pixels (width × height) Dimensione originale in pixel (larghezza × altezza) - + Processing size in pixels (width × height), scale factor Dimensione dell'elaborazione in pixel (larghezza × altezza), fattore di scala @@ -837,7 +837,7 @@ Per impostazione predefinita, utilizza il profilo incorporato. Tuttavia, è poss File TIFF non supportati - + The following 5-channel TIFF files are not supported: I seguenti file TIFF a 5 canali non sono supportati: @@ -864,27 +864,27 @@ Per impostazione predefinita, utilizza il profilo incorporato. Tuttavia, è poss split - + Number of boxes removed by the OCR model / total boxes Numero di caselle rimosse dal modello OCR / caselle totali - + Number of boxes that failed to generate a mask / total boxes Numero di caselle che non hanno generato una maschera / caselle in totale - + Number of boxes that were perfectly masked / total boxes Numero di caselle perfettamente mascherate / totale delle caselle - + Number of boxes that were denoised / total boxes Numero di caselle che sono state denoizzate / caselle totali - + Number of boxes that were inpainted Numero di caselle dipinte @@ -1032,14 +1032,14 @@ Per impostazione predefinita, utilizza il profilo incorporato. Tuttavia, è poss Richiesta impossibile - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Il denoising è disabilitato nel profilo corrente, questo output non può essere'generato. Attivare il denoising nelle impostazioni del profilo e riprovare. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. L'inpainting è disabilitato nel profilo corrente, questo output non può'essere generato. @@ -1061,7 +1061,7 @@ Attivare l'inpainting nelle impostazioni del profilo e riprovare. Esportazione fallita - + Failed to export image: Impossibile esportare l'immagine: @@ -1104,7 +1104,7 @@ Attivare l'inpainting nelle impostazioni del profilo e riprovare. risultato - + results risultati @@ -1119,17 +1119,17 @@ Attivare l'inpainting nelle impostazioni del profilo e riprovare. nuova immagine - + new images nuove immagini - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Corrispondenza tra {num_images} {plural_images} e i risultati OCR. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} OCR orfani {plural_result} andranno persi. @@ -1228,12 +1228,12 @@ Attivare l'inpainting nelle impostazioni del profilo e riprovare. Nota: il nome "{name}" è stato nascosto - + Corrupted log session Sessione di log corrotta - + Current session Sessione in corso @@ -1248,7 +1248,7 @@ Attivare l'inpainting nelle impostazioni del profilo e riprovare. Errore - + Errors Errori @@ -1258,7 +1258,7 @@ Attivare l'inpainting nelle impostazioni del profilo e riprovare. Critico - + Criticals Criticità @@ -1274,49 +1274,49 @@ Attivare l'inpainting nelle impostazioni del profilo e riprovare. MainWindow - + Open Config Statusbar button Configurazione aperta - + Open Log Statusbar button Registro aperto - + CUDA Enabled CUDA abilitato - + Select Output Directory Selezionare la directory di uscita - + Aborting... Interruzione... - + Loading OCR model... Caricamento del modello OCR... - + Loaded OCR model. Modello OCR caricato. - + Error Errore - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1325,7 +1325,7 @@ Continue anyway? Continuare comunque? - + Delete Models Cancellare i modelli @@ -1335,23 +1335,23 @@ Continuare comunque? Siete sicuri di voler eliminare i modelli di apprendimento automatico? Questo renderà impossibile la pulizia e l'OCR finché non verranno scaricati di nuovo. - + Encountered error: Errore riscontrato: - + MPS Enabled MPS abilitato - - + + Inpainting Model Update Aggiornamento del Modello di Inpainting - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1365,148 +1365,148 @@ Passare al nuovo modello? Il vecchio modello è stato mantenuto. Per eseguire l'aggiornamento, seleziona "Aiuto" e poi "Eliminare i Modelli di Apprendimento Automatico" dalla barra dei menu. - + Layout Reset Reset del layout - + The window layout has been reset. The default layout will be used next time. Il layout della finestra è stato ripristinato. La prossima volta verrà utilizzato il layout predefinito. - + Multiple Instances Istanze multiple - + Found process ID in lock file: Trovato l'ID del processo nel file di blocco: - + Failed to delete models. Errore nell'eliminazione dei modelli. - + Delete Failed Eliminazione non riuscita - + Models Deleted Modelli eliminati - + The models were deleted. Would you like to download them again? I modelli sono stati cancellati. Volete scaricarli di nuovo? - + Not Supported Non supportato - + Unsupported system or package format Formato del sistema o del pacchetto non supportato - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Riavvio richiesto - + The language has been changed. Please restart the application for the changes to take effect. La lingua è stata modificata. Per rendere effettive le modifiche, riavviare l'applicazione. - - + + System Language Lingua del sistema - + Import Profile Profilo di importazione - + Profile Files (*.conf) File di profilo (*.conf) - + Profile Imported Profilo Importato - + Import Error Errore di importazione - + Failed to Delete Cancellazione non riuscita - + The default profile cannot be deleted. Il profilo predefinito non può essere eliminato. - + Delete Profile Cancellare il profilo - + Are you sure you want to delete the profile {profile_name}? Sei sicuro di voler cancellare il profilo {profile_name}? - + Delete Error Cancellare l'errore - + Failed to delete the profile. Non è stato possibile eliminare il profilo. @@ -1518,217 +1518,217 @@ Switching profiles will discard changes to the current profile. Se si cambia profilo, le modifiche apportate al profilo corrente vengono eliminate. - + Unsaved changes Modifiche non salvate - - + + Load Error Errore di caricamento - - + + Failed to load profile: Impossibile caricare il profilo: - + Tesseract OCR is not installed or not found Tesseract OCR non è installato o non viene trovato - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Non è possibile'utilizzare Tesseract per eseguire l'OCR. Ritorno a manga-ocr. Vedere le istruzioni per installare correttamente Tesseract <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">qui</a> o continuare a usare il modello predefinito.</html> - + Unsupported Language Lingua non supportata - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. La lingua '{lang_name}' non è supportata da nessuno dei motori OCR attuali. Per sapere come aggiungere il supporto per altre lingue, consultare la documentazione online. - + Profile Created Profilo creato - + Create Error Crea errore - + Failed to save profile to {profile_path} Impossibile salvare il profilo su {profile_path} - - + + Save Error Salva errore - + Failed to save profile. Impossibile salvare il profilo. - + Profile saved to {profile_path} Profilo salvato su {profile_path} - + Failed to save config. Impossibile salvare la configurazione. - + Failed to save the new profile to the configuration file. Non è stato possibile salvare il nuovo profilo nel file di configurazione. - - + + No Files Nessun file - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. Nessun file da elaborare. È possibile aggiungere file trascinandoli al centro della finestra o tramite la barra dei menu: File -> Aggiungi file o Aggiungi cartella. - + No Outputs Nessuna uscita - + No outputs were requested. Please select at least one output before cleaning. Non è stata richiesta alcuna uscita. Selezionare almeno un'uscita prima della pulizia. - - + + cleaned pulito - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. Nessun file da elaborare. Per modificare un vecchio file di output OCR, è necessario prima caricare (una o più) delle immagini a cui corrisponde. - + Open OCR Output File Aprire il file di output OCR - + OCR Output Files (*.txt *.csv) File di output OCR (*.txt *.csv) - + Parse Error Errore di analisi - + Failed to parse the OCR output file. Impossibile analizzare il file di output OCR. - + File Exists Il file esiste - + The file '{output_path}' already exists. Overwrite? Il file '{output_path}' esiste già. Sovrascrivere? - + Export Images Esportazione di immagini - + Would you like to export the cleaned images? Volete esportare le immagini pulite? - + Export OCR Results Esportazione dei risultati OCR - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Modifiche non applicate - + You have un-applied changes to the profile. Continue anyway? Le modifiche al profilo non sono state applicate. Continuare comunque? - + Would you like to export the OCR results? Si desidera esportare i risultati dell'OCR? - + Processing Finished Elaborazione terminata - + Finished processing all files. Finito di elaborare tutti i file. - + Processing Aborted Elaborazione interrotta - + Processing aborted. Elaborazione interrotta. - + Processing Error Errore di elaborazione - + Encountered an error while processing files. Si è verificato un errore durante l'elaborazione dei file. @@ -1739,7 +1739,7 @@ Per sapere come aggiungere il supporto per altre lingue, consultare la documenta - + Profile Profilo @@ -1754,360 +1754,365 @@ Per sapere come aggiungere il supporto per altre lingue, consultare la documenta Applicare - - + + Save Salvare - + Reset All Resettare tutto - + Drag and Drop Images or Folders Here Trascina qui le immagini o le cartelle - + Images Immagini - - + + File File - + Size Dimensioni - + Processing Size The size that the picture is shrunk down to for processing. Elaborazione delle dimensioni - + File Size Dimensione del file - + Color Mode Like RGB, grayscale etc. Modalità colore - + Analytics Can also call this statistics. Analisi - + Process Processo - + Cleaning Pulizia - + OCR OCR - + Edit Existing Output Modifica dell'output esistente - + Abort Interruzione - + Start Inizio - + Output Uscita - + Cleaned Image Immagine pulita - + Mask Maschera - + Isolated Text Testo isolato - - + + Review Output Controllo Uscita - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>Se selezionata, al termine del processo si apre una finestra di controllo. I risultati sono anche pre-visualizzabili nella vista dei dettagli di ciascuna immagine'prima di procedere al salvataggio su disco.</p></body></html> - + Save Output Salvare Uscita - + Output Directory: Directory di Uscita: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. È possibile utilizzare un percorso relativo per creare una sottocartella nella posizione originale dell'immagine'oppure utilizzare un percorso assoluto. - + Plain Text Testo normale - + When checked, a review window will open upon process completion. Quando è selezionata, al termine del processo si apre una finestra di revisione. - + Output File: File di uscita: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Inserire un percorso assoluto o relativo con un nome di file. L'output di tutte le immagini viene scritto nello stesso file.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. testo_rilevato.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Generazione: - + Current Step: Passo attuale: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Immagini elaborate: - + Set Default Impostazione predefinita - + Settings Impostazioni - + Theme As in color theme Tema - + Language Lingua - + Help Aiuto - + Add Files... Aggiungi file... - + Remove All Files Rimuovere tutti i file - + Remove File Rimuovi file - + Help Translate Panel Cleaner Aiuto per la traduzione di Panel Cleaner - + Report an Issue... Segnala un problema... - + Simulate Exception Simulare l'eccezione - + Add File Browser Extension... Aggiungi estensione browser file... - + Show Supported OCR Languages Mostra le lingue OCR supportate - - + + Reset Window Layout Preferences Resettare le preferenze di layout della finestra - + Show Out Of Memory Warnings Mostra avvisi di memoria esaurita - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New Nuovo - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>Se si seleziona, gli output vengono salvati su disco. Altrimenti, è possibile visualizzarli in anteprima solo nella vista Dettagli immagine o nella Revisione output, se attivata. </p><p>Se il profilo rimane invariato dopo la pulizia senza questa opzione, è possibile esportarlo rapidamente eseguendo nuovamente la pulizia con questa opzione attivata.</p></body></html> - + CSV Spreadsheet Foglio di calcolo CSV - + Delete Cancellare - + Import... Importazione... - + Online Documentation Documentazione online - + About Panel Cleaner Informazioni su Panel Cleaner - + View License Visualizza la licenza - + Save as... Salva con nome... - + System Sistema - + Dark Scuro - + Light Luce - + Add Folder... Aggiungi cartella... - + Delete Machine Learning Models Eliminare i modelli di apprendimento automatico - + Download Machine Learning Models Scarica i modelli di Machine Learning - + Donate Donare @@ -2139,19 +2144,19 @@ Per sapere come aggiungere il supporto per altre lingue, consultare la documenta Modello OCR: - + Already downloaded Già scaricato - + Text Detector model (CUDA) Modello di rivelatore di testo (CUDA) - + Text Detector model (CPU) Modello di rilevatore di testo (CPU) @@ -2170,7 +2175,7 @@ Per sapere come aggiungere il supporto per altre lingue, consultare la documenta Errori di download dell'OCR - + ETA estimated time of completion Tempo stimato di arrivo @@ -2748,7 +2753,7 @@ Per sapere come aggiungere il supporto per altre lingue, consultare la documenta The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2829,73 +2834,73 @@ Per sapere come aggiungere il supporto per altre lingue, consultare la documenta Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input L'immagine originale con il ridimensionamento applicato (se necessario). - + Input Step name in the image details view Entrata - + Input Output name in the image details view Entrata - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask La maschera grezza generata dall'IA. - + Text Detection Step name in the image details view Rilevamento del testo - + AI Mask Output name in the image details view Maschera AI - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes I dati della casella non filtrati generati dall'IA. - + Raw Boxes Output name in the image details view Caselle grezze - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes I contorni delle caselle di testo trovate dall'IA. - + Preprocessor Step name in the image details view Preprocessore - + Initial Boxes Output name in the image details view Caselle iniziali - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2903,181 +2908,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Verde: caselle iniziali. Rosso: caselle estese. Viola: caselle unite (finali). Blu: caselle di riferimento per il denoising. - + Final Boxes Output name in the image details view Caselle Finali - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask La maschera delle caselle unite. - + Masker Step name in the image details view Mascheratore - + Box Mask Output name in the image details view Casella Maschera - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask La maschera di rilevamento del testo grezzo con tutto ciò che si trova al di fuori della maschera della casella tagliata. - + Cut Mask Output name in the image details view Maschera di taglio - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers Le diverse fasi di crescita intorno alla maschera di taglio vengono visualizzate con colori diversi. - + Mask Layers Output name in the image details view Livelli di maschera - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality La deviazione standard (σ) e lo spessore del contorno (in pixel) di ciascuna maschera migliore scelta, se presente. σ più basso è meglio, da perfetto (viola) a fallito (rosso). - + Fitment Quality Output name in the image details view Qualità di montaggio - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay L'immagine di entrata con la maschera finale sovrapposta a colori. - + Mask Overlay Output name in the image details view Sovrapposizione di maschere - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask La collezione di maschere per ogni bolla che si adatta meglio. - + Final Mask Output name in the image details view Maschera finale - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text Il livello di testo isolato dall'immagine di input. - + Isolated Text Output name in the image details view Testo isolato - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output L'immagine di input con la maschera finale applicata. - + Masked Output Output name in the image details view Uscita mascherata - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask Le maschere che richiedono il denoising, da sovrapporre alla maschera finale durante l'esportazione. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Maschera Denoise - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output L'immagine di ingresso con la maschera con denoising applicato. - + Denoised Output Output name in the image details view Uscita con Denoiser - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask Le sezioni dipinte intorno al testo sono state pulite male, se non del tutto. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Maschera dipinta - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output L'immagine di ingresso: pulita, con denoiser (se abilitato) e con inpainter. - + Inpainted Output Output name in the image details view Uscita con Inpainting - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3086,1141 +3091,1141 @@ Verde: caselle iniziali. Rosso: caselle estese. Viola: caselle unite (finali). B Profile - + General Profile section title Generale - + About this profile: Profile comment in General Informazioni su questo profilo: - + Notes Profile entry in General Note - + Preferred file type to save the cleaned image as. Profile comment in General Tipo di file preferito per salvare l'immagine pulita. - + Preferred File Type Profile entry in General Tipo di file preferito - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Tipo di file preferito per salvare la maschera. Sono supportati solo i formati immagine che consentono la trasparenza. - + Preferred Mask File Type Profile entry in General Tipo di file della maschera preferito - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General I seguenti sono gli obiettivi inferiore e superiore per l'altezza dell'immagine di entrata. L'immagine viene ridimensionata solo per rientrare nell'intervallo, preferendo fattori numerici interi per ridurre al minimo l'impatto sulla qualità dell'immagine. Le immagini più piccole di uno dei due target rimarranno invariate. È possibile disattivare questa funzione impostando uno o entrambi i valori a un valore inferiore o uguale a 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General È utile per accelerare in modo significativo l'elaborazione di immagini di grandi dimensioni. Inoltre, dato che le altre opzioni che si basano sulle dimensioni dei pixel dipendono dalle dimensioni, questa opzione aiuta a normalizzare i risultati tra immagini di dimensioni diverse. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General L'immagine viene ridimensionata, elaborata e quindi solo la maschera viene ridimensionata. Ciò significa che l'output pulito utilizzerà ancora l'immagine originale non ridimensionata per evitare qualsiasi perdita di qualità. Per determinare il fattore di scala viene utilizzata solo l'altezza dell'immagine, preservando il rapporto d'aspetto e ignorando la larghezza individuale di un'immagine, in modo che il fattore rimanga coerente se una delle pagine è doppia. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General Ad esempio, per un target inferiore di 1000 e un target superiore di 2000, un'immagine di dimensioni 5000x7000 (w, h) pixel verrà ridimensionata di un fattore 4, in modo da avere le dimensioni di 1250x1750 pixel durante l'elaborazione. - + Input Height Lower Target Profile entry in General Altezza di entrata Obiettivo inferiore - + Input Height Upper Target Profile entry in General Altezza di entrata Obiettivo superiore - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Numero massimo di threads da utilizzare per l'esportazione delle immagini. Si può lasciare non specificato per utilizzare tutti i threads disponibili. Abbassare questo valore se si riscontrano problemi di memoria, che si manifestano come crash casuali. - + Max Threads Export Profile entry in General Esportazione di Threads massimi - + Text Detector Profile section title Rilevatore di testo - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Percorso del modello di rilevamento del testo; lasciare vuoto per usare il modello incorporato. Si consiglia di sovrascriverlo solo se la versione scaricata automaticamente è più vecchia dell'ultima release. Il percorso deve puntare direttamente al file comictextdetector.pt (CUDA) o comictextdetector.pt.onnx (CPU). È possibile scaricare versioni precedenti del modello <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">qui.</a> - + Model Path Profile entry in Text Detector Percorso del modello - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Numero di modelli da eseguire in parallelo. È utile se si dispone di una quantità di RAM (o VRAM con CUDA) sufficiente per eseguire più modelli contemporaneamente. Questo, ovviamente, aumenta la velocità del processo, ma può anche mandare in crash il computer se si sovrastima l'hardware. Si consiglia di utilizzare 1 modello per 2 GB di memoria disponibile, oppure 1 modello per 50 immagini. Nota: questa regola viene ignorata se si elaborano meno di 50 file, poiché al di sotto di questa soglia non vale la pena avviare più modelli. Attenzione: Questo può causare l'instabilità del programma, usatelo a vostro rischio e pericolo. <br>NON segnalate problemi relativi a questa impostazione, perché'dipende interamente dall'hardware! - + Concurrent Models Profile entry in Text Detector Modelli concorrenti - + Preprocessor Profile section title Preprocessore - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Le dimensioni delle caselle sono indicate nel numero totale di pixel, quindi una casella di 200x200 pixel ha una dimensione di 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Dimensione minima di una casella per conservarla. - + Box Min Size Profile entry in Preprocessor Casella Dimensione minima - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Dimensione minima di una casella con "linguaggio sconosciuto" per mantenerlo. Questo linguaggio viene tipicamente assegnato ai loghi e ad altri elementi di testo che si comportano male. - + Suspicious Box Min Size Profile entry in Preprocessor Casella Sospetta Dimensione minima - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor La sovrapposizione minima tra due riquadri per unirli. Questa percentuale (0-100) indica quanta parte della casella più piccola deve trovarsi all'interno di quella più grande per essere unita. Un valore più alto richiederà una sovrapposizione maggiore per unire le caselle. - + Box Overlap Threshold Profile entry in Preprocessor Soglia di sovrapposizione della casella - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Se utilizzare l'OCR per individuare le caselle che'non meritano di essere pulite, come quelle che contengono solo numeri o simboli. - + OCR Enabled Profile entry in Preprocessor OCR abilitato - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Se si desidera utilizzare Tesseract per eseguire operazioni OCR.<br> Se è selezionata, Tesseract OCR può essere utilizzato per l'estrazione del testo, se disponibile.<br> Se non è selezionata, viene sempre utilizzato il modello OCR integrato (manga-ocr), che è più adatto per il testo giapponese verticale. - + OCR Use Tesseract Profile entry in Preprocessor OCR Utilizzare Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor La lingua da utilizzare per le attività OCR. Il rilevatore di testo è in grado di rilevare il giapponese e l'inglese; per tutto il resto, selezionare esplicitamente una lingua. Il rilevamento per casella mantiene ciò che il rilevatore di testo ha rilevato, mentre per pagina assegnerà la lingua più importante all'intera pagina.<br> Nota: solo il giapponese è supportato immediatamente, tutto il resto richiede l'abilitazione di Tesseract e l'installazione dei language pack associati. - + OCR Language Profile entry in Preprocessor Lingua OCR - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifica quale motore utilizzare per l'esecuzione dell'OCR.<br> - auto: Seleziona automaticamente il motore OCR in base alla lingua rilevata di ciascun blocco di testo all'interno dell'immagine. Utilizza Manga Ocr per il testo giapponese, Tesseract per il testo inglese o sconosciuto.<br> - mangaocr: Impone a Panel Cleaner di utilizzare il modello manga-ocr integrato per tutte le operazioni di riconoscimento del testo. È più adatto per il testo giapponese verticale.<br> - tesseract: consente a Panel Cleaner di utilizzare l'OCR Tesseract per tutte le attività di riconoscimento del testo. Si tratta di un'opzione versatile che supporta l'inglese e diverse altre lingue. - + OCR Engine Profile entry in Preprocessor Motore OCR - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Definisce l'ordine di lettura per l'elaborazione e l'ordinamento delle caselle di testo dell'intera pagina, non dei singoli blocchi di testo. Questa impostazione globale influenza il modo in cui le caselle di testo vengono ordinate e presentate per l'ulteriore elaborazione.<br> - auto: Rileva l'ordine di lettura in base alla lingua rilevata di ciascun blocco di testo all'interno della pagina.<br> - manga: Ordine da destra a sinistra, dall'alto in basso. Adatto ai manga giapponesi.<br> - fumetto: Ordine da sinistra a destra, dall'alto in basso. Adatto a fumetti e testi occidentali.<br> Scegliere in base al layout predominante del contenuto. - + Reading Order Profile entry in Preprocessor Ordine di lettura - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Dimensione massima di una casella su cui eseguire l'OCR. Queste caselle inutili sono di solito piccole e l'OCR è lento, quindi si usa questo valore come limite. - + OCR Max Size Profile entry in Preprocessor OCR Dimensione massima - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Modello Regex da confrontare con i risultati dell'OCR. Tutto ciò che corrisponde a questo modello viene scartato.<br> Nota: il modello OCR restituisce caratteri a tutta larghezza, quindi questo modello dovrebbe corrispondere. - + OCR Blacklist Pattern Profile entry in Preprocessor Schema di lista nera OCR - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor Il modello MangaOCR è in grado di gestire solo il testo giapponese, quindi quando è abilitata l'opzione forzata del modello, scarterà le caselle che non sono'sicure che siano giapponesi. A volte, numeri o altri simboli riducono l'efficacia, facendo sì che la lingua rilevata sia sconosciuta. Se l'opzione forzata è disabilitata, questi non verranno scartati. Tutto ciò che è riconosciuto con sicurezza come lingua diversa sarà scartato indipendentemente da questa impostazione.<br> Nota: questa impostazione è rilevante solo quando ocr_language è impostato su rilevamento per casella o pagina. - + OCR Strict Language Profile entry in Preprocessor Lingua rigorosa OCR - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding da aggiungere a ciascun lato di una casella. Questo viene aggiunto alle caselle iniziali create dal rilevatore di testo AI. Queste caselle sono visualizzate in verde. - + Box Padding Initial Profile entry in Preprocessor Padding della casella iniziale - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding da aggiungere al lato destro di una casella. Questa estensione aiuta a coprire il testo in rubino che fluttua a destra del testo verticale. - + Box Right Padding Initial Profile entry in Preprocessor Padding casella destra iniziale - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding da aggiungere a ciascun lato di una casella. Questo viene aggiunto a una serie estesa di caselle, utilizzate per eliminare i falsi positivi dalla maschera del rilevatore di testo AI'. Queste caselle sono visualizzate in viola. - + Box Padding Extended Profile entry in Preprocessor Padding della scatola estesa - + Box Right Padding Extended Profile entry in Preprocessor Padding casella destra estesa - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding da aggiungere a ciascun lato di una casella. Questo viene aggiunto alle caselle di riferimento utilizzate per campionare l'immagine originale durante l'analisi della maschera da utilizzare. Queste caselle sono visualizzate in blu. - + Box Reference Padding Profile entry in Preprocessor Padding del riferimento della casella - + Masker Profile section title Mascheratore - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Numero massimo di thread da utilizzare per l'esportazione delle immagini. Si può lasciare non specificato per utilizzare tutti i thread disponibili. Abbassare questo valore se si riscontrano problemi di memoria, che si manifestano come crash casuali. - + Max Threads Profile entry in Masker Threads massimi - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Numero di pixel per far crescere la maschera a ogni passo. Questo ingrandisce il contorno della maschera, quindi valori più piccoli saranno più precisi ma più lenti. - + Mask Growth Step Pixels Profile entry in Masker Maschera di crescita Fase Pixel - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Numero di passi con cui far crescere la maschera. Un numero più alto creerà maschere sempre più grandi, limitate in ultima analisi dalle dimensioni del riquadro di riferimento. - + Mask Growth Steps Profile entry in Masker Fasi di crescita della maschera - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Spessore minimo di una maschera. Questo è come il passo di crescita della prima maschera'e i restanti seguiranno il passo di crescita della maschera Pixel. In questo modo, è possibile avere una dimensione di passo ridotta per garantire la precisione, ma evitare maschere molto sottili che potrebbero essere utilizzate per pulire il testo solo circondato da un contorno, quando l'inpittura sarebbe'la scelta migliore. - + Min Mask Thickness Profile entry in Masker Spessore minimo della maschera - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combinare gli output in un unico file di progetto come layer diversi. Formati attualmente supportati: Photoshop PSD.<br> - nessuno: Ogni immagine e maschera vengono salvate come file di base.<br> - psd-per-image: Le immagini e le maschere vengono salvate insieme in un file PSD per ogni immagine di input.<br> - psd-bulk: Tutte le immagini e le maschere vengono salvate insieme in un unico file PSD, raggruppato per immagine di input. - + Layered Export Profile entry in General Esportazione in Layers - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Se consentire o meno le maschere colorate. Quando è -true-, il mascheratore permetterà alle maschere di usare qualsiasi colore, non solo il bianco, il nero o il grigio. - + Allow Colored Masks Profile entry in Masker Consentire le maschere colorate - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Soglia massima per un pixel da considerare off-white. Il colore mediano lungo il bordo di una maschera potrebbe non essere bianco puro, quindi per evitare lievi tonalità di off-white su uno sfondo bianco puro, tutto ciò che è più chiaro di questo valore di soglia verrà arrotondato al bianco puro. Value range: black (0) to pure white (255).']".". - + Off White Max Threshold Profile entry in Masker Bianco sporco Soglia massima - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker La deviazione standard di una maschera rappresenta la variazione di colore lungo il bordo della maschera. Per questo, solo la singola linea di pixel lungo il bordo esterno di una maschera viene campionata dall'immagine originale. Una deviazione bassa significa che la maschera si trova in un colore solido, il che significa che non'interseca alcun testo o altro oggetto. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Miglioramento minimo della deviazione standard della maschera per continuare a farla crescere. L'impostazione di un valore più alto richiede un miglioramento maggiore per considerare una maschera più grande, per dare una preferenza alle maschere più piccole. - + Mask Improvement Threshold Profile entry in Masker Soglia di miglioramento della maschera - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Se utilizzare l'algoritmo di selezione rapida delle maschere. Quando è vero, l'algoritmo di selezione della maschera sceglierà la prima maschera perfetta, se ne viene trovata una in anticipo. Questo è più veloce, ma potrebbe non trovare la maschera migliore, se una leggermente più grande sarebbe stata migliore. - + Mask Selection Fast Profile entry in Masker Selezione della maschera Veloce - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Numero massimo di threads da utilizzare per l'esportazione delle immagini. Si può lasciare non specificato per utilizzare tutti i threads disponibili. Abbassare questo valore se si riscontrano problemi di memoria, che si manifestano come crash casuali. - + Max Threads Profile entry in Denoiser Threads massimi - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter FUNZIONE SPERIMENTALE: se si trovano impostazioni predefinite migliori, aprire un problema su github per condividere i miglioramenti con tutti. Si noti che l'inpainting non è'abilitato per impostazione predefinita. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter L'inpainting è l'uso dell'apprendimento automatico per sostituire il contenuto di un'immagine in base all'ambiente circostante.'Per le maschere che non hanno potuto essere pulite bene (o affatto), si può ricorrere all'inpainting. Per distinguerlo dal denoising, l'inpainting è destinato a maschere significativamente peggiori che avevano un adattamento stretto. Tutte le maschere che sono state denoising non saranno'inpaintate. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Poiché in alcuni casi questa fase può fornire risultati scadenti, è possibile disabilitarla. Deselezionare per disabilitare l'inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting abilitato - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter Il raggio massimo di una maschera su cui eseguire l'inpainting. Le maschere più grandi saranno lasciate così come sono, perché se il margine è così grande, è probabile che la maschera sia già abbastanza buona. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Raggio massimo della maschera - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter Il raggio minimo intorno a una maschera da dipingere. Viene aggiunto alla dimensione ottimale della maschera per garantire che l'inpittura copra l'intera maschera. - + Min Inpainting Radius Profile entry in Inpainter Raggio minimo di Impainting - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter Per le maschere che si sono rivelate molto più difficili da pulire, ovvero con un'elevata deviazione standard, aumentare il raggio dell'inpainting per coprire una parte maggiore della maschera. Questo margine aggiuntivo viene aggiunto al raggio minimo di inpainting e viene calcolato come: moltiplicatore del raggio di inpainting per la deviazione standard della maschera - + Inpainting Radius Multiplier Profile entry in Inpainter Moltiplicatore del raggio di Impainting - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter Il raggio massimo intorno a una maschera da dipingere. Questo limita le dimensioni che l'inpainting può raggiungere, per evitare che copra troppo l'immagine, se si usa un moltiplicatore di raggio elevato. - + Max Inpainting Radius Profile entry in Inpainter Raggio massimo di Impainting - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter Dopo l'inpainting, tagliare il risultato dall'immagine originale per evitare che l'inpainting influisca sul resto dell'immagine. In questo modo si garantisce che l'immagine originale venga preservata il più possibile. Questo raggio viene aggiunto intorno al raggio di inpainting finale, a causa del modello di inpainting che modifica alcuni pixel al di fuori della sua regione dedicata. - + Inpainting Isolation Radius Profile entry in Inpainter Raggio di isolamento dell 'Impainting - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Rileva per casella - + Detect per page Language option for OCR Rileva per pagina - + Japanese Language option for OCR Giapponese - + English Language option for OCR Inglese - + Korean Language option for OCR Coreano - + Korean (vertical) Language option for OCR Coreano (verticale) - + Chinese - Simplified Language option for OCR Cinese semplificato - + Chinese - Traditional Language option for OCR Cinese tradizionale - + Albanian Language option for OCR Albanese - + Arabic Language option for OCR Arabo - + Azerbaijani Language option for OCR Azerbaigiano - + Azerbaijani - Cyrilic Language option for OCR Azerbaigiano - Cirillico - + Bengali Language option for OCR Bengalese - + Bulgarian Language option for OCR Bulgaro - + Burmese Language option for OCR Birmano - + Catalan; Valencian Language option for OCR Catalano; Valenciano - + Croatian Language option for OCR Croato - + Czech Language option for OCR Ceco - + Danish Language option for OCR Danese - + Dutch; Flemish Language option for OCR Olandese; Fiammingo - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estone - + Finnish Filipino Language option for OCR Finlandese - + French Language option for OCR Francese - + Georgian Language option for OCR Georgiano - + German Language option for OCR Tedesco - + Greek Language option for OCR Greco - + Hebrew Language option for OCR Ebraico - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Ungherese - + Indonesian Language option for OCR Indonesiano - + Italian Language option for OCR Italiano - + Kazakh Language option for OCR Kazako - + Latin Language option for OCR Latino - + Lithuanian Language option for OCR Lituano - + Luxembourgish Language option for OCR Lussemburghese - + Malay Language option for OCR Malese - + Mongolian Language option for OCR Mongolo - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norvegese - + Persian Language option for OCR Persiano - + Polish Language option for OCR Polacco - + Portuguese Language option for OCR Portoghese - + Romanian; Moldavian Language option for OCR Rumeno; moldavo - + Russian Language option for OCR Russo - + Serbian Language option for OCR Serbo - + Serbian - Latin Language option for OCR Serbo - Latino - + Slovak Language option for OCR Slovacco - + Slovenian Language option for OCR Sloveno - + Spanish; Castilian Language option for OCR Spagnolo; Castigliano - + Swedish Language option for OCR Svedese - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thailandese - + Turkish Language option for OCR Turco - + Ukrainian Language option for OCR Ucraino - + Vietnamese Language option for OCR Vietnamita - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker La deviazione standard massima di una maschera da considerare. Un valore elevato significa una maggiore tolleranza per la maschera che interseca il testo o altri oggetti, il che'non è una buona maschera, in quanto richiederà comunque una ridipintura. Impostando questo valore a 0, si possono ottenere solo maschere perfette, il che è consigliabile per le immagini ad alta risoluzione. - + Mask Max Standard Deviation Profile entry in Masker Maschera Deviazione standard massima - + Color to use for the debug mask. Profile comment in Masker Colore da usare per la maschera di debug. - + Debug Mask Color Profile entry in Masker Colore della maschera di debug - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser Quando un po' di testo è difficile da mascherare, la fase di pulizia ha probabilmente dovuto scegliere una maschera piccola, che lascia dietro di sé molti artefatti jpeg, se erano intorno al testo. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser Ciò significa che il colore dei pixel intorno al bordo di una maschera non è'uniforme, il che viene quantificato come deviazione standard. Il denoiser può puntare a maschere con una deviazione standard minima e ridurre il rumore dell'area proprio intorno alla maschera. In questo modo si conservano i dettagli nel resto dell'immagine, ma si rimuovono gli artefatti proprio nel punto in cui si trovava il testo. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Poiché si tratta di un passaggio facoltativo e potrebbe essere superfluo per le immagini ad alta risoluzione che non'soffrono di artefatti jpeg, può essere disabilitato qui. Impostare su False per disabilitare il denoising. - + Denoising Enabled Profile entry in Denoiser Denoising abilitato - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser La deviazione standard minima dei colori intorno al bordo di una determinata maschera per eseguire il denoising sulla regione intorno alla maschera. - + Noise Min Standard Deviation Profile entry in Denoiser Rumore Deviazione standard minima - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Nota: se l'inpainting è abilitato, la deviazione std minima dell'inpainting fungerà da massimo per questo, subito dopo la maschera verrà consegnata all'inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser Lo spessore di un contorno per il denoising intorno a una maschera. - + Noise Outline Size Profile entry in Denoiser Rumore Dimensioni del contorno - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Sfumare i bordi della maschera di copertura denoisata di questa quantità per fondere uniformemente le parti con denoiser nel resto dell'immagine. - + Noise Fade Radius Profile entry in Denoiser Raggio di dissolvenza del rumore - + Set to True to support colored images. Profile comment in Denoiser Impostare su True per supportare le immagini colorate. - + Colored Images Profile entry in Denoiser Immagini colorate - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser Quanto forte è il denoising dell'immagine. Valori più alti significano più denoising. - + Filter Strength Profile entry in Denoiser Forza del filtro - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser Quanto forte è il denoising dei canali colore, rispetto alla luminosità. Valori più alti significano più denoising. - + Color Filter Strength Profile entry in Denoiser Forza del filtro colore - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Dimensione in pixel della patch del modello utilizzata per calcolare i pesi. Dovrebbe essere dispari. - + Template Window Size Profile entry in Denoiser Dimensioni della finestra del modello - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Dimensione in pixel della finestra utilizzata per calcolare la media ponderata per un determinato pixel. Deve essere dispari. - + Search Window Size Profile entry in Denoiser Dimensione della finestra di ricerca @@ -4247,7 +4252,7 @@ Verde: caselle iniziali. Rosso: caselle estese. Viola: caselle unite (finali). B px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. Tutti i nuclei @@ -4664,14 +4669,14 @@ Verde: caselle iniziali. Rosso: caselle estese. Viola: caselle unite (finali). B ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>Il percorso del modello di Text Detector {value} non esiste, si torna al modello predefinito. È possibile scaricare manualmente il modello da <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">qui</a> o continuare a usare il modello predefinito.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From c1c5662836aa9c01c7f90b6b0585c9114a8140a9 Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:16 +0100 Subject: [PATCH 11/18] New translations panelcleaner_source.ts (Japanese) --- translations/PanelCleaner_ja_JP.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_ja_JP.ts b/translations/PanelCleaner_ja_JP.ts index 894a147..4893ef2 100644 --- a/translations/PanelCleaner_ja_JP.ts +++ b/translations/PanelCleaner_ja_JP.ts @@ -4,7 +4,7 @@ - + default reserved profile name @@ -20,54 +20,54 @@ Profile {profile_name} created. - + Profile name cannot be empty. Profile name cannot be empty. - + Profile name already in use. Profile name already in use. - + Profile name is reserved. Profile name is reserved. - - - - + + + + Unknown Unknown - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Grayscale - + 1-bit Color mode 1-bit - + Palette Color mode Palette @@ -78,45 +78,45 @@ Error downloading file from url: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Error downloading file from url: {url} Failed to save the file to {save_path} - + Error downloading file from url: {url} The file content is different from expected. Error downloading file from url: {url} The file content is different from expected. + - - - + + + - - - + + N/A N/A - + OCR Analytics OCR Analytics - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) @@ -155,66 +155,66 @@ Removed bubbles: - + Small boxes Small boxes - + Removed boxes Removed boxes - - + + Failed Failed - + Perfect Perfect - + Total Total - + Mask Fitment Analytics Mask Fitment Analytics - + Total Total - + Total boxes Total boxes - + Masks succeeded Masks succeeded - + Masks failed Masks failed - + Perfect masks Perfect masks - + Average border deviation Average border deviation @@ -233,7 +233,7 @@ Mask usage by thickness (in pixels): Mask ({thickness}px) - + Box mask Box mask @@ -254,27 +254,27 @@ Pages with failures / total: - + Denoising Analytics Denoising Analytics - + Total masks Total masks - + Masks denoised Masks denoised - + Minimum deviation to denoise Minimum deviation to denoise - + Maximum allowed deviation Maximum allowed deviation @@ -293,27 +293,27 @@ Pages with failures / total: Denoised - + Inpainting Analytics Inpainting Analytics - + Inpainting performed Inpainting performed - + Average thickness Average thickness - + Minimum thickness Minimum thickness - + Maximum thickness Maximum thickness @@ -372,7 +372,7 @@ Pages with failures / total: The file format was not recognized. - + The CSV file must start with a header row, followed by data rows. The CSV file must start with a header row, followed by data rows. @@ -402,7 +402,7 @@ Pages with failures / total: Multiple images match the file path: {path} - + Error, no version found. Error, no version found. @@ -412,7 +412,7 @@ Pages with failures / total: CUDA Error - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pages with failures / total: Inpainting model not found. - + filename filename - + startx startx - + starty starty - + endx endx - + endy endy - + text text - + Saved detected text to {output_file} Saved detected text to {output_file} - + Failed to write detected text to {output_file} Failed to write detected text to {output_file} - + Save Failed Save Failed - + Failed to write detected text to file. Failed to write detected text to file. @@ -534,7 +534,7 @@ Pages with failures / total: - + Page layered export @@ -620,7 +620,7 @@ An AI-powered tool to clean manga panels. ColorButton - + Select Color Select Color @@ -666,7 +666,7 @@ An AI-powered tool to clean manga panels. Unsupported system - + This system is not supported. This system is not supported. @@ -681,12 +681,12 @@ An AI-powered tool to clean manga panels. Uninstall for Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. The extension will be installed at {path}. This doesn't require root permissions. - + The extension file will be deleted. This doesn't require root permissions. The extension file will be deleted. This doesn't require root permissions. @@ -701,17 +701,17 @@ An AI-powered tool to clean manga panels. Uninstall for Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. The extension will be installed in the Windows registry.This requires admin permissions. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. The extension will be uninstalled from the Windows registry.This requires admin permissions. - + Dolphin extension installed successfully. Dolphin extension installed successfully. @@ -721,7 +721,7 @@ An AI-powered tool to clean manga panels. Installation successful - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Dolphin extension installed successfully. @@ -801,12 +801,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) Original size in pixels (width × height) - + Processing size in pixels (width × height), scale factor Processing size in pixels (width × height), scale factor @@ -832,7 +832,7 @@ By default, this uses the built-in profile. However, you can change this default Unsupported TIFF files - + The following 5-channel TIFF files are not supported: The following 5-channel TIFF files are not supported: @@ -859,27 +859,27 @@ By default, this uses the built-in profile. However, you can change this default split - + Number of boxes removed by the OCR model / total boxes Number of boxes removed by the OCR model / total boxes - + Number of boxes that failed to generate a mask / total boxes Number of boxes that failed to generate a mask / total boxes - + Number of boxes that were perfectly masked / total boxes Number of boxes that were perfectly masked / total boxes - + Number of boxes that were denoised / total boxes Number of boxes that were denoised / total boxes - + Number of boxes that were inpainted Number of boxes that were inpainted @@ -1027,14 +1027,14 @@ By default, this uses the built-in profile. However, you can change this default Impossible Request - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting is disabled in the current profile, this output can't be generated. @@ -1056,7 +1056,7 @@ Please enable inpainting in the profile settings and try again. Export failed - + Failed to export image: Failed to export image: @@ -1099,7 +1099,7 @@ Please enable inpainting in the profile settings and try again. result - + results results @@ -1114,17 +1114,17 @@ Please enable inpainting in the profile settings and try again. new image - + new images new images - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. @@ -1223,12 +1223,12 @@ Please enable inpainting in the profile settings and try again. Note: Name "{name}" was hidden - + Corrupted log session Corrupted log session - + Current session Current session @@ -1243,7 +1243,7 @@ Please enable inpainting in the profile settings and try again. Error - + Errors Errors @@ -1253,7 +1253,7 @@ Please enable inpainting in the profile settings and try again. Critical - + Criticals Criticals @@ -1269,49 +1269,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button Open Config - + Open Log Statusbar button Open Log - + CUDA Enabled CUDA Enabled - + Select Output Directory Select Output Directory - + Aborting... Aborting... - + Loading OCR model... Loading OCR model... - + Loaded OCR model. Loaded OCR model. - + Error Error - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1320,7 +1320,7 @@ Continue anyway? Continue anyway? - + Delete Models Delete Models @@ -1330,23 +1330,23 @@ Continue anyway? Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. - + Encountered error: Encountered error: - + MPS Enabled MPS Enabled - - + + Inpainting Model Update Inpainting Model Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1360,148 +1360,148 @@ Switch to the new model? Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. - + Layout Reset Layout Reset - + The window layout has been reset. The default layout will be used next time. The window layout has been reset. The default layout will be used next time. - + Multiple Instances Multiple Instances - + Found process ID in lock file: Found process ID in lock file: - + Failed to delete models. Failed to delete models. - + Delete Failed Delete Failed - + Models Deleted Models Deleted - + The models were deleted. Would you like to download them again? The models were deleted. Would you like to download them again? - + Not Supported Not Supported - + Unsupported system or package format Unsupported system or package format - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Restart Required - + The language has been changed. Please restart the application for the changes to take effect. The language has been changed. Please restart the application for the changes to take effect. - - + + System Language System Language - + Import Profile Import Profile - + Profile Files (*.conf) Profile Files (*.conf) - + Profile Imported Profile Imported - + Import Error Import Error - + Failed to Delete Failed to Delete - + The default profile cannot be deleted. The default profile cannot be deleted. - + Delete Profile Delete Profile - + Are you sure you want to delete the profile {profile_name}? Are you sure you want to delete the profile {profile_name}? - + Delete Error Delete Error - + Failed to delete the profile. Failed to delete the profile. @@ -1513,217 +1513,217 @@ Switching profiles will discard changes to the current profile. Switching profiles will discard changes to the current profile. - + Unsaved changes Unsaved changes - - + + Load Error Load Error - - + + Failed to load profile: Failed to load profile: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Profile Created - + Create Error Create Error - + Failed to save profile to {profile_path} Failed to save profile to {profile_path} - - + + Save Error Save Error - + Failed to save profile. Failed to save profile. - + Profile saved to {profile_path} Profile saved to {profile_path} - + Failed to save config. Failed to save config. - + Failed to save the new profile to the configuration file. Failed to save the new profile to the configuration file. - - + + No Files No Files - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. - + No Outputs No Outputs - + No outputs were requested. Please select at least one output before cleaning. No outputs were requested. Please select at least one output before cleaning. - - + + cleaned cleaned - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists File Exists - + The file '{output_path}' already exists. Overwrite? The file '{output_path}' already exists. Overwrite? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Processing Finished - + Finished processing all files. Finished processing all files. - + Processing Aborted Processing Aborted - + Processing aborted. Processing aborted. - + Processing Error Processing Error - + Encountered an error while processing files. Encountered an error while processing files. @@ -1734,7 +1734,7 @@ Check the online documentation for how to add support for more languages. - + Profile Profile @@ -1749,360 +1749,365 @@ Check the online documentation for how to add support for more languages.Apply - - + + Save Save - + Reset All Reset All - + Drag and Drop Images or Folders Here Drag and Drop Images or Folders Here - + Images Images - - + + File File - + Size Size - + Processing Size The size that the picture is shrunk down to for processing. Processing Size - + File Size File Size - + Color Mode Like RGB, grayscale etc. Color Mode - + Analytics Can also call this statistics. Analytics - + Process Process - + Cleaning Cleaning - + OCR OCR - + Edit Existing Output Edit Existing Output - + Abort Abort - + Start Start - + Output Output - + Cleaned Image Cleaned Image - + Mask Mask - + Isolated Text Isolated Text - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Save Output - + Output Directory: Output Directory: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. You can use a relative path to create a subfolder at the image's original location, or use an absolute path. - + Plain Text Plain Text - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Output File: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Generating: - + Current Step: Current Step: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Images Processed: - + Set Default Set Default - + Settings Settings - + Theme As in color theme Theme - + Language Language - + Help Help - + Add Files... Add Files... - + Remove All Files Remove All Files - + Remove File Remove File - + Help Translate Panel Cleaner Help Translate Panel Cleaner - + Report an Issue... Report an Issue... - + Simulate Exception Simulate Exception - + Add File Browser Extension... Add File Browser Extension... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New New - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Delete - + Import... Import... - + Online Documentation Online Documentation - + About Panel Cleaner About Panel Cleaner - + View License View License - + Save as... Save as... - + System System - + Dark Dark - + Light Light - + Add Folder... Add Folder... - + Delete Machine Learning Models Delete Machine Learning Models - + Download Machine Learning Models Download Machine Learning Models - + Donate Donate @@ -2134,19 +2139,19 @@ Check the online documentation for how to add support for more languages.OCR model: - + Already downloaded Already downloaded - + Text Detector model (CUDA) Text Detector model (CUDA) - + Text Detector model (CPU) Text Detector model (CPU) @@ -2165,7 +2170,7 @@ Check the online documentation for how to add support for more languages.OCR download errors - + ETA estimated time of completion ETA @@ -2743,7 +2748,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2824,73 +2829,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input The original image with scaling applied (if needed). - + Input Step name in the image details view Input - + Input Output name in the image details view Input - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask The rough mask generated by the AI. - + Text Detection Step name in the image details view Text Detection - + AI Mask Output name in the image details view AI Mask - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes The unfiltered box data generated by the AI. - + Raw Boxes Output name in the image details view Raw Boxes - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes The outlines of the text boxes the AI found. - + Preprocessor Step name in the image details view Preprocessor - + Initial Boxes Output name in the image details view Initial Boxes - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2898,181 +2903,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. - + Final Boxes Output name in the image details view Final Boxes - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask The mask of the merged boxes. - + Masker Step name in the image details view Masker - + Box Mask Output name in the image details view Box Mask - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask The rough text detection mask with everything outside the box mask cut out. - + Cut Mask Output name in the image details view Cut Mask - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers The different steps of growth around the cut mask displayed in different colors. - + Mask Layers Output name in the image details view Mask Layers - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). - + Fitment Quality Output name in the image details view Fitment Quality - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay The input image with the final mask overlaid in color. - + Mask Overlay Output name in the image details view Mask Overlay - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask The collection of masks for each bubble that fit best. - + Final Mask Output name in the image details view Final Mask - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text The text layer isolated from the input image. - + Isolated Text Output name in the image details view Isolated Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output The input image with the final mask applied. - + Masked Output Output name in the image details view Masked Output - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask The masks that required denoising, to be overlaid on the final mask when exporting. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Denoise Mask - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output The input image with the denoised mask applied. - + Denoised Output Output name in the image details view Denoised Output - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask The inpainted sections around the text that was poorly cleaned, if at all. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Inpainted Mask - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output The input image: cleaned, denoised (if enabled), and inpainted. - + Inpainted Output Output name in the image details view Inpainted Output - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3081,1141 +3086,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title General - + About this profile: Profile comment in General About this profile: - + Notes Profile entry in General Notes - + Preferred file type to save the cleaned image as. Profile comment in General Preferred file type to save the cleaned image as. - + Preferred File Type Profile entry in General Preferred File Type - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Preferred file type to save the mask as. Only image formats that allow for transparency are supported. - + Preferred Mask File Type Profile entry in General Preferred Mask File Type - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. - + Input Height Lower Target Profile entry in General Input Height Lower Target - + Input Height Upper Target Profile entry in General Input Height Upper Target - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Text Detector - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> - + Model Path Profile entry in Text Detector Model Path - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! - + Concurrent Models Profile entry in Text Detector Concurrent Models - + Preprocessor Profile section title Preprocessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Minimum size of any box to keep it. - + Box Min Size Profile entry in Preprocessor Box Min Size - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. - + Suspicious Box Min Size Profile entry in Preprocessor Suspicious Box Min Size - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. - + Box Overlap Threshold Profile entry in Preprocessor Box Overlap Threshold - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. - + OCR Enabled Profile entry in Preprocessor OCR Enabled - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. - + OCR Use Tesseract Profile entry in Preprocessor OCR Use Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. - + OCR Engine Profile entry in Preprocessor OCR Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. - + Reading Order Profile entry in Preprocessor Reading Order - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. - + OCR Max Size Profile entry in Preprocessor OCR Max Size - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR Blacklist Pattern - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor OCR Strict Language - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. - + Box Padding Initial Profile entry in Preprocessor Box Padding Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. - + Box Right Padding Initial Profile entry in Preprocessor Box Right Padding Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Box Right Padding Extended - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. - + Box Reference Padding Profile entry in Preprocessor Box Reference Padding - + Masker Profile section title Masker - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. - + Mask Growth Step Pixels Profile entry in Masker Mask Growth Step Pixels - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. - + Mask Growth Steps Profile entry in Masker Mask Growth Steps - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. - + Min Mask Thickness Profile entry in Masker Min Mask Thickness - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Off White Max Threshold - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. - + Mask Improvement Threshold Profile entry in Masker Mask Improvement Threshold - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. - + Mask Selection Fast Profile entry in Masker Mask Selection Fast - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting Enabled - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. - + Min Inpainting Radius Profile entry in Inpainter Min Inpainting Radius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation - + Inpainting Radius Multiplier Profile entry in Inpainter Inpainting Radius Multiplier - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. - + Max Inpainting Radius Profile entry in Inpainter Max Inpainting Radius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. - + Inpainting Isolation Radius Profile entry in Inpainter Inpainting Isolation Radius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. - + Mask Max Standard Deviation Profile entry in Masker Mask Max Standard Deviation - + Color to use for the debug mask. Profile comment in Masker Color to use for the debug mask. - + Debug Mask Color Profile entry in Masker Debug Mask Color - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. - + Denoising Enabled Profile entry in Denoiser Denoising Enabled - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. - + Noise Min Standard Deviation Profile entry in Denoiser Noise Min Standard Deviation - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser The thickness of an outline to denoise around a mask. - + Noise Outline Size Profile entry in Denoiser Noise Outline Size - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. - + Noise Fade Radius Profile entry in Denoiser Noise Fade Radius - + Set to True to support colored images. Profile comment in Denoiser Set to True to support colored images. - + Colored Images Profile entry in Denoiser Colored Images - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the image. Higher values mean more denoising. - + Filter Strength Profile entry in Denoiser Filter Strength - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. - + Color Filter Strength Profile entry in Denoiser Color Filter Strength - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Size in pixels of the template patch that is used to compute weights. Should be odd. - + Template Window Size Profile entry in Denoiser Template Window Size - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. - + Search Window Size Profile entry in Denoiser Search Window Size @@ -4242,7 +4247,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4659,14 +4664,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From 2f8bd9d2b15ac44b3f7c1518cb2b425d73f71ec4 Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:18 +0100 Subject: [PATCH 12/18] New translations panelcleaner_source.ts (Korean) --- translations/PanelCleaner_ko_KR.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_ko_KR.ts b/translations/PanelCleaner_ko_KR.ts index 4760e7f..c2d239f 100644 --- a/translations/PanelCleaner_ko_KR.ts +++ b/translations/PanelCleaner_ko_KR.ts @@ -4,7 +4,7 @@ - + default reserved profile name 기본 @@ -20,54 +20,54 @@ 프로파일 {profile_name}이(가) 생성됨. - + Profile name cannot be empty. 프로파일 이름은 비어 있을 수 없습니다. - + Profile name already in use. 이미 사용 중인 프로파일 이름입니다. - + Profile name is reserved. 예약된 프로파일 이름입니다. - - - - + + + + Unknown 알 수 없음 - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode 회색조 - + 1-bit Color mode 1비트 - + Palette Color mode 팔레트 @@ -78,45 +78,45 @@ 다음 URL에서 파일을 다운로드하는데 오류가 발생했습니다: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} 다음 URL에서 파일을 다운로드하는데 오류가 발생했습니다: {url} 파일을 {save_path}에 저장하는데 실패했습니다 - + Error downloading file from url: {url} The file content is different from expected. 다음 URL에서 파일을 다운로드하는데 오류가 발생했습니다: {url} 파일의 내용이 예상과 다릅니다. + - - - + + + - - - + + N/A N/A - + OCR Analytics OCR 분석 - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) 글자 영역 개수: {num_boxes} | 작은 영역 개수: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) 제거된 영역 수: {num_removed} (전체 대비 {removed_box_ratio}, 작은 영역 개수 중 {removed_among_small_ratio}) @@ -155,66 +155,66 @@ Removed bubbles: - + Small boxes 작은 영역 - + Removed boxes 제거된 글자 영역 - - + + Failed 실패 - + Perfect 완벽 - + Total 전체 - + Mask Fitment Analytics 마스크 적합성 분석 - + Total 전체 - + Total boxes 전체 글자 영역 개수 - + Masks succeeded 마스크 생성 성공 - + Masks failed 마스크 생성 실패 - + Perfect masks 완벽한 마스크 - + Average border deviation 평균 경계 편차 @@ -233,7 +233,7 @@ Mask usage by thickness (in pixels): 마스크 ({thickness}px) - + Box mask 영역 마스크 @@ -254,27 +254,27 @@ Pages with failures / total: - + Denoising Analytics 노이즈 감소 분석 - + Total masks 전체 마스크 개수 - + Masks denoised 노이즈 감소 처리된 마스크 - + Minimum deviation to denoise 노이즈 감소 최소 편차 - + Maximum allowed deviation 최대 허용 편차 @@ -293,27 +293,27 @@ Pages with failures / total: 노이즈 감소됨 - + Inpainting Analytics 인페인팅 분석 - + Inpainting performed 인페인팅 처리한 개수 - + Average thickness 평균 두께 - + Minimum thickness 최소 두께 - + Maximum thickness 최대 두께 @@ -372,7 +372,7 @@ Pages with failures / total: 인식할 수 없는 파일 형식입니다. - + The CSV file must start with a header row, followed by data rows. CSV 파일은 반드시 헤더 행으로 시작하고, 이어서 데이터 행이 나와야 합니다. @@ -402,7 +402,7 @@ Pages with failures / total: 여러 이미지가 해당 파일 경로와 일치합니다: {path} - + Error, no version found. 오류. 버전을 찾을 수 없음. @@ -412,7 +412,7 @@ Pages with failures / total: CUDA 오류 - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>현재 사용 중인 GPU는 필요한 CUDA 연산을 지원하지 않습니다.<br><br>현재 설치된 torch 및 torchvision을 제거하고, 대신 CPU 버전(또는 다른 CUDA 버전)을 설치해 보세요.<br>자세한 설치 방법은 다음 페이지에서 확인할 수 있습니다:<br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>"Compute Platform" 부분에서 사용 가능한 CUDA 버전을 확인하세요.<br><br>현재 사용 중인 CUDA 버전: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pages with failures / total: 인페인팅 모델을 찾을 수 없음. - + filename filename - + startx startx - + starty starty - + endx endx - + endy endy - + text text - + Saved detected text to {output_file} 감지된 글자들을 다음 경로에 저장했습니다: {output_file} - + Failed to write detected text to {output_file} 감지된 글자들을 다음 경로에 저장하는 데 실패했습니다: {output_file} - + Save Failed 저장 실패 - + Failed to write detected text to file. 감지된 글자들을 파일에 쓰는 데 실패했습니다. @@ -534,7 +534,7 @@ Pages with failures / total: - + Page layered export @@ -620,7 +620,7 @@ An AI-powered tool to clean manga panels. ColorButton - + Select Color 색상 선택 @@ -666,7 +666,7 @@ An AI-powered tool to clean manga panels. 미지원 시스템 - + This system is not supported. 이 시스템은 지원하지 않습니다. @@ -681,12 +681,12 @@ An AI-powered tool to clean manga panels. 돌핀 삭제 - + The extension will be installed at {path}. This doesn't require root permissions. 확장 기능이 {path}에 설치될 예정입니다. 이 작업은 루트 권한이 필요하지 않습니다. - + The extension file will be deleted. This doesn't require root permissions. 확장 기능이 삭제될 예정입니다. 이 작업은 루트 권한이 필요하지 않습니다. @@ -701,17 +701,17 @@ An AI-powered tool to clean manga panels. 윈도우 파일 탐색기를 위해 삭제 - + The extension will be installed in the Windows registry.This requires admin permissions. 확장 기능이 윈도우 레지스트리에 설치될 예정입니다. 이 작업은 관리자 권한이 필요합니다. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. 확장 기능이 윈도우 레지스트리에서 삭제될 예정입니다. 이 작업은 관리자 권한이 필요합니다. - + Dolphin extension installed successfully. 돌핀 확장 기능이 설치되었습니다. @@ -721,7 +721,7 @@ An AI-powered tool to clean manga panels. 설치 성공 - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. 돌핀 확장 기능이 성공적으로 설치되었습니다. @@ -801,12 +801,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) 원본 픽셀 크기(가로 x 세로) - + Processing size in pixels (width × height), scale factor 처리 시 픽셀 크기(가로 x 세로), 스케일 배수 @@ -832,7 +832,7 @@ By default, this uses the built-in profile. However, you can change this default 지원하지 않는 TIFF 파일 - + The following 5-channel TIFF files are not supported: 다음 5채널 TIFF 파일은 지원하지 않습니다: @@ -859,27 +859,27 @@ By default, this uses the built-in profile. However, you can change this default 분할 - + Number of boxes removed by the OCR model / total boxes OCR 모델에 의해 지워진 글자 영역 수 / 전체 글자 영역 수 - + Number of boxes that failed to generate a mask / total boxes 마스크 생성에 실패한 영역 수 / 전체 글자 영역 수 - + Number of boxes that were perfectly masked / total boxes 성공적으로 생성된 마스크 개수 / 전체 글자 영역 수 - + Number of boxes that were denoised / total boxes 노이즈 감소된 영역 수 / 전체 글자 영역 수 - + Number of boxes that were inpainted 인페인팅 된 영역 수 @@ -1027,14 +1027,14 @@ By default, this uses the built-in profile. However, you can change this default 불가능한 요청 - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. 현재 프로파일에서 노이즈 감소가 비활성화되어 있어 이 출력을 생성할 수 없습니다. 프로파일에서 노이즈 감소를 활성화하고 다시 시도하세요. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. 현재 프로파일에서 인페인팅이 비활성화되어 있어 이 출력을 생성할 수 없습니다. @@ -1056,7 +1056,7 @@ Please enable inpainting in the profile settings and try again. 내보내기 실패 - + Failed to export image: 이미지 내보내기 실패: @@ -1099,7 +1099,7 @@ Please enable inpainting in the profile settings and try again. 결과 - + results 결과 @@ -1114,17 +1114,17 @@ Please enable inpainting in the profile settings and try again. 새 이미지 - + new images 새 이미지 - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. OCR 결과에 {num_images}개의 이미지가 매칭되었습니다. 새 이미지 {num_images_unassigned}개가 남았습니다. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} 개의 분리된 OCR 결과가 손실됩니다. @@ -1223,12 +1223,12 @@ Please enable inpainting in the profile settings and try again. 참고: 로그에서 사용자 이름 "{name}"은(는) 가려졌습니다 - + Corrupted log session 중단된 로그 세션 - + Current session 현재 세션 @@ -1243,7 +1243,7 @@ Please enable inpainting in the profile settings and try again. 개 오류 - + Errors 개 오류 @@ -1253,7 +1253,7 @@ Please enable inpainting in the profile settings and try again. 개 심각한 오류 - + Criticals 개 심각한 오류 @@ -1269,49 +1269,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button 설정 열기 - + Open Log Statusbar button 로그 열기 - + CUDA Enabled CUDA 활성화됨 - + Select Output Directory 저장 디렉토리 설정 - + Aborting... 중지중... - + Loading OCR model... OCR 모델 불러오는 중... - + Loaded OCR model. OCR 모델 불러옴. - + Error 오류 - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1320,7 +1320,7 @@ Continue anyway? 계속하시겠습니까? - + Delete Models 모델 삭제 @@ -1330,23 +1330,23 @@ Continue anyway? 머신 러닝 모델을 삭제하시겠습니까? 다시 다운로드할 때까지 글자 지우기 및 OCR이 불가능해집니다. - + Encountered error: 오류 발생: - + MPS Enabled MPS 활성화됨 - - + + Inpainting Model Update 인페인팅 모델 업데이트 - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1360,148 +1360,148 @@ Switch to the new model? 오래된 모델을 유지합니다. 업데이트하려면 메뉴에서 "도움말"을 선택한 다음 "머신 러닝 모델 삭제"를 선택하세요. - + Layout Reset 레이아웃 초기화 - + The window layout has been reset. The default layout will be used next time. 화면의 레이아웃을 초기화합니다. 다음부터 기본값이 적용됩니다. - + Multiple Instances 다중 인스턴스 - + Found process ID in lock file: lock 파일에서 찾은 프로세스 ID: - + Failed to delete models. 모델 삭제 실패. - + Delete Failed 삭제 실패 - + Models Deleted 모델 삭제됨 - + The models were deleted. Would you like to download them again? 모델을 삭제했습니다. 다시 다운로드하시겠습니까? - + Not Supported 지원되지 않음 - + Unsupported system or package format 지원되지 않는 시스템 또는 패키지 형식 - + Shutdown 컴퓨터 종료 - + After processing, the system will shut down. 처리가 끝난 후 컴퓨터를 종료합니다. - + You have review options enabled, these will need to be manually closed before the action can start. 검토가 활성화되어 있으므로 작업을 시작하기 전에 수동으로 닫아야 합니다. - + You have disabled writing output. 쓰기 출력을 비활성화했습니다. - + Shutting down system... 시스템 종료... - + Running post action "{action}"... 후속 작업 "{action}" 실행... - + Restart Required 재시작 필요 - + The language has been changed. Please restart the application for the changes to take effect. 언어가 변경되었습니다. 변경 사항을 적용하려면 프로그램을 종료 후 다시 실행해 주세요. - - + + System Language 시스템 언어 - + Import Profile 프로파일 가져오기 - + Profile Files (*.conf) 프로파일 파일 (*.conf) - + Profile Imported 프로파일 가져옴 - + Import Error 가져오기 오류 - + Failed to Delete 삭제 실패 - + The default profile cannot be deleted. 기본 프로파일은 삭제할 수 없습니다. - + Delete Profile 프로파일 삭제 - + Are you sure you want to delete the profile {profile_name}? 정말로 프로파일 {profile_name}을(를) 삭제하시겠습니까? - + Delete Error 삭제 오류 - + Failed to delete the profile. 프로파일 삭제 실패. @@ -1513,217 +1513,217 @@ Switching profiles will discard changes to the current profile. 프로파일을 전환하면 현재 변경 사항이 없어집니다. - + Unsaved changes 저장되지 않은 변경 사항 - - + + Load Error 불러오기 오류 - - + + Failed to load profile: 프로파일 불러오기 실패: - + Tesseract OCR is not installed or not found 테서렉트 OCR이 설치되어 있지 않거나 찾을 수 없습니다 - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>테서렉트를 사용할 수 없어 OCR을 할 수 없습니다. manga-ocr로 되돌립니다. 테서렉트를 올바르게 설치하기 위해 <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">이 문서</a>를 읽거나 기본 모델을 씁니다.</html> - + Unsupported Language 지원되지 않는 언어 - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. 언어 '{lang_name}'는 모든 현재 OCR 모델에서 지원되지 않습니다. 더 많은 언어를 지원하는 방법은 온라인 문서를 확인하세요. - + Profile Created 프로파일 생성됨 - + Create Error 생성 오류 - + Failed to save profile to {profile_path} 프로파일을 다음 경로에 저장하는 데 실패했습니다: {profile_path} - - + + Save Error 저장 오류 - + Failed to save profile. 프로파일 저장 실패. - + Profile saved to {profile_path} 프로파일이 다음 경로에 저장됐습니다: {profile_path} - + Failed to save config. 설정 저장 실패. - + Failed to save the new profile to the configuration file. 새 프로파일을 설정 파일로 저장하는 데 실패했습니다. - - + + No Files 파일 없음 - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. 처리할 파일이 없습니다. 화면에 드래그 앤 드롭하거나 메뉴를 통해 파일을 추가할 수 있습니다: 파일 -> 파일 추가 또는 폴더 추가. - + No Outputs 출력 없음 - + No outputs were requested. Please select at least one output before cleaning. 저장할 결과물이 없습니다. 저장할 것을 한 개 이상 선택해 주세요. - - + + cleaned 정리된이미지 - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. 처리할 파일이 없습니다. 이전 OCR 결과 파일을 편집하려면 먼저 해당 파일에 해당하는 이미지를 로드해야 합니다(하나 이상). - + Open OCR Output File OCR 결과 파일 열기 - + OCR Output Files (*.txt *.csv) OCR 결과 파일 (*.txt *.csv) - + Parse Error 파싱 오류 - + Failed to parse the OCR output file. OCR 결과 파일을 파싱하는 데 실패했습니다. - + File Exists 파일 존재함 - + The file '{output_path}' already exists. Overwrite? 피일 '{output_path}'은 이미 존재합니다. 덮어쓰시겠습니까? - + Export Images 이미지 내보내기 - + Would you like to export the cleaned images? 정리된 이미지들을 내보내겠습니까? - + Export OCR Results OCR 결과 내보내기 - + After processing, the "{action}" action will be executed. 처리가 완료된 후 작업 "{action}"이(가) 실행됩니다. - - + + Post action canceled. 후속 작업 취소됨. - + Un-Applied Changes 적용되지 않은 변경 사항 - + You have un-applied changes to the profile. Continue anyway? 프로파일에 적용되지 않은 변경 사항이 있습니다. 그래도 계속하시겠습니까? - + Would you like to export the OCR results? OCR 결과를 내보내겠습니까? - + Processing Finished 처리 완료 - + Finished processing all files. 모든 파일 처리 완료. - + Processing Aborted 처리 중단됨 - + Processing aborted. 처리가 중단되었습니다. - + Processing Error 처리 오류 - + Encountered an error while processing files. 처리 중 오류가 발생했습니다. @@ -1734,7 +1734,7 @@ Check the online documentation for how to add support for more languages. - + Profile 프로파일 @@ -1749,360 +1749,365 @@ Check the online documentation for how to add support for more languages.적용 - - + + Save 저장 - + Reset All 모두 초기화 - + Drag and Drop Images or Folders Here 폴더나 이미지를 여기로 끌고 오세요 - + Images 이미지 - - + + File 파일 - + Size 해상도 - + Processing Size The size that the picture is shrunk down to for processing. 처리 시의 크기 - + File Size 파일 크기 - + Color Mode Like RGB, grayscale etc. 색상 모드 - + Analytics Can also call this statistics. 처리 현황 - + Process 작업 - + Cleaning 글자 지우기 - + OCR OCR(글자 인식) - + Edit Existing Output 이전 인식 결과 수정 - + Abort 중단 - + Start 시작 - + Output 내보내기 - + Cleaned Image 지워진 이미지 - + Mask 마스크 - + Isolated Text 분리된 글자 - - + + Review Output 결과물 검토 - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>체크하면 처리가 완료된 후 검토 화면이 열립니다. 결과물을 저장하기 전에 각 이미지의 세부 정보를 미리 볼 수 있습니다.</p></body></html> - + Save Output 결과물 저장 - + Output Directory: 저장할 폴더의 경로: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. 상대 경로 또는 절대 경로를 사용할 수 있으며, 상대 경로의 경우 기준은 이미지의 위치입니다. - + Plain Text 일반 텍스트 - + When checked, a review window will open upon process completion. 체크하면 처리가 완료된 후 검토 화면이 열립니다. - + Output File: 파일 저장 경로: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>파일 이름을 포함하는 상대 경로 형식을 쓰거나 절대 경로를 사용할 수 있습니다.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action 작업 취소 - + Close 닫기 - + Generating: This is present progressive, as in "[currently] Generating:" 처리 중: - + Current Step: 현재 단계: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. 처리된 이미지: - + Set Default 사용할 프로파일 - + Settings 설정 - + Theme As in color theme 테마 - + Language 언어 - + Help 도움말 - + Add Files... 파일 추가... - + Remove All Files 모든 파일 제거 - + Remove File 파일 제거 - + Help Translate Panel Cleaner 패널 클리너 번역 도와주기 - + Report an Issue... 이슈 제보... - + Simulate Exception 예외 시뮬레이션 - + Add File Browser Extension... 파일 관리자 확장 기능 추가... - + Show Supported OCR Languages 지원하는 OCR 언어 표시 - - + + Reset Window Layout Preferences 화면 레이아웃 설정 초기화 - + Show Out Of Memory Warnings 메모리 부족 경고 표시 - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions 후속 작업 - + Configure Post-Run Actions... 후속 작업 구성하기... - + New 새로 만들기 - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>체크하면 결과물이 저장됩니다. 그렇지 않으면 이미지 세부 정보 보기 또는 검토 작업(설정한 경우)만 할 수 있습니다. </p><p>이 옵션을 체크하지 않은 상태에서 작업 후 프로파일이 변경되지 않았다면 이 옵션을 체크하고 지우기 작업을 다시 실행하면 빠르게 내보낼 수 있습니다.</p></body></html> - + CSV Spreadsheet CSV - + Delete 삭제 - + Import... 가져오기... - + Online Documentation 온라인 문서 - + About Panel Cleaner 패널 클리너에 대해 - + View License 라이선스 보기 - + Save as... 다른 이름으로 저장... - + System 시스템 설정에 따름 - + Dark 다크 - + Light 라이트 - + Add Folder... 폴더 추가... - + Delete Machine Learning Models 머신 러닝 모델 삭제 - + Download Machine Learning Models 머신 러닝 모델 다운로드 - + Donate 후원 @@ -2134,19 +2139,19 @@ Check the online documentation for how to add support for more languages.OCR 모델: - + Already downloaded 이미 다운로드됨 - + Text Detector model (CUDA) 글자 감지 모델 (CUDA) - + Text Detector model (CPU) 글자 감지 모델 (CPU) @@ -2165,7 +2170,7 @@ Check the online documentation for how to add support for more languages.OCR 다운로드 오류 - + ETA estimated time of completion 남은 예상 시간 @@ -2743,7 +2748,7 @@ Check the online documentation for how to add support for more languages.이름 "{shutdown_command}"은 예약되어 있으므로 사용할 수 없습니다. - + The Flatpak sandbox will likely prevent actions from working. Flatpak 샌드박스 환경에서는 작업이 제대로 실행되지 않을 수 있습니다. @@ -2824,73 +2829,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input 원본 이미지입니다. 필요한 경우 크기가 변경될 수 있습니다. - + Input Step name in the image details view 입력 - + Input Output name in the image details view 입력 - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask AI가 생성한 마스크 초안입니다. - + Text Detection Step name in the image details view 글자 감지 - + AI Mask Output name in the image details view AI 마스크 - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes AI가 생성한 필터링되지 않은 글자 영역입니다. - + Raw Boxes Output name in the image details view 원시 영역 - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes AI가 찾은 글자 영역입니다. - + Preprocessor Step name in the image details view 전처리기 - + Initial Boxes Output name in the image details view 초기 글자 영역 - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2898,181 +2903,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r 녹색: 초기 글자 영역. 빨간색: 확장 글자 영역. 보라색: 병합된(최종) 글자 영역. 파란색: 노이즈 감소를 위한 참조 영역. - + Final Boxes Output name in the image details view 최종 글자 영역 - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask 병합된 글자 영역들의 마스크입니다. - + Masker Step name in the image details view 마스크 - + Box Mask Output name in the image details view 글자 영역 마스크 - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask 글자 마스크의 초안입니다. - + Cut Mask Output name in the image details view 글자 마스크 - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers 마스크 주변의 각 확장 단계를 여러 가지 색으로 표시합니다. - + Mask Layers Output name in the image details view 마스크 확장 단계 - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality 선택된 최적 마스크 각각의 표준 편차(σ)와 두께(픽셀 단위)를 표시합니다. 표준 편차값이 낮을수록 더 좋은 마스크이며, 완벽한 경우는 보라색, 마스크 생성에 실패한 경우는 빨간색으로 표시됩니다. - + Fitment Quality Output name in the image details view 최적 마스크 - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay 최종 마스크를 색상으로 오버레이한 이미지입니다. - + Mask Overlay Output name in the image details view 마스크 오버레이 - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask 각 말풍선의 최적 마스크들입니다. - + Final Mask Output name in the image details view 최종 마스크 - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text 이미지에서 분리된 글자 레이어입니다. - + Isolated Text Output name in the image details view 글자 추출 - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output 최종 마스크가 적용된 이미지입니다. - + Masked Output Output name in the image details view 마스크 처리된 결과 - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask 노이즈 감소 작업이 필요했던 마스크로, 내보낼 때 최종 마스크 위에 오버레이 됩니다. - + Denoiser Step name in the image details view 노이즈 감소 - + Denoise Mask Output name in the image details view 노이즈 감소된 마스크 - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output 노이즈 감소된 마스크가 적용된 원본 이미지입니다. - + Denoised Output Output name in the image details view 노이즈 감소 결과 - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask 제대로 지워지지 않은 글자 영역 주변에 적용된 인페인팅 결과입니다. 경우에 따라 아무것도 없을 수 있습니다. - + Inpainter Step name in the image details view 인페인팅 - + Inpainted Mask Output name in the image details view 인페인팅 된 마스크 - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output 지우기, 노이즈 감소, 인페인팅이 모두 적용된 최종 이미지입니다. - + Inpainted Output Output name in the image details view 인페인팅 된 결과 - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3081,1141 +3086,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title 일반 - + About this profile: Profile comment in General 프로파일 설명: - + Notes Profile entry in General 참고 - + Preferred file type to save the cleaned image as. Profile comment in General 완성된 이미지를 저장할 형식. - + Preferred File Type Profile entry in General 저장 형식 - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General 마스크 이미지를 저장할 형식. 투명도를 지원하는 형식만 가능합니다. - + Preferred Mask File Type Profile entry in General 마스크 저장 형식 - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General 주어지는 이미지의 높이에 대한 하한 및 상한 설정입니다. 이미지 품질에 미치는 영향을 최소화하기 위해 정수값을 쓰는 것이 좋으며, 범위 내에 맞게만 축소됩니다. 하나 혹은 두 값을 0 이하로 설정하여 이 기능을 비활성화할 수 있습니다. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General 이 옵션은 큰 이미지의 처리 속도를 높이는데 유용합니다. 또한 픽셀 크기에 의존하는 옵션들의 결과를 정규화하는 데 도움이 됩니다. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General 이미지가 축소되어 처리된 다음, 마스크만 다시 확대됩니다. 즉, 처리된 결과물에는 품질 손실 방지를 위해 크기 조정이 되지 않은 원본 이미지가 그대로 사용됩니다. 이미지의 높이만 배율을 결정하는 데 사용되므로 이미지 전체의 비율은 유지되고 각 너비는 무시되므로 이미지 중 하나가 두 페이지를 담고 있는 경우 배율이 일관되게 유지됩니다. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General 예를 들어 하한이 1000이고 상한이 2000인 경우 가로x세로가 5000x7000픽셀인 이미지는 처리 중 1250x1750픽셀로 4배 축소됩니다. - + Input Height Lower Target Profile entry in General 높이 하한 - + Input Height Upper Target Profile entry in General 높이 상한 - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General 긴 콘텐츠를 개별 페이지로 분할합니다. 활성화하면 전체 콘텐츠를 원하는 높이로 맞추기 위해 찌그러뜨리지 않고 개별 이미지로 분할하는데, 이때 높이는 "분할 목표 높이" ± ("분할 여백 허용범위" * 2)의 값 내에서 정해집니다. - + Split Long Strips Profile entry in General 긴 콘텐츠 분할 허용 - + Preferred height to split long strips at. Profile comment in General 분할되기를 원하는 높이 값. - + Preferred Split Height Profile entry in General 분할 목표 높이 - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General 긴 이미지 분할 시 사용될 여백 허용 범위입니다. 분할 목표 높이 값과 실제 분할 결과의 높이 간의 최대 차이입니다. 알고리즘은 이미지가 컷 중간에서 잘리는 것을 방지하기 위해 이 여백 범위 내에서 최적의 분할 지점을 결정합니다. - + Split Tolerance Margin Profile entry in General 분할 여백 허용범위 - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General 긴 이미지를 분할할 때 사용할 기준 종횡비로, 이미지의 너비와 높이의 비율입니다. 이미지의 종횡비가 이 값보다 작으면 긴 이미지로 간주합니다.<br>예시: 0.3은 이미지의 너비가 높이의 0.3배임을 의미합니다. - + Long Strip Aspect Ratio Profile entry in General 긴 이미지로 간주할 비율 - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General 긴 이미지를 다시 단일 이미지로 병합합니다. 이 옵션을 활성화하면 분할된 각 이미지가 내보낼 때 다시 하나의 이미지로 합쳐집니다. - + Merge After Split Profile entry in General 분할 후 다시 합치기 - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General 이미지 내보내기에 사용할 최대 스레드 수입니다. 이 값을 지정하지 않으면 사용할 수 있는 모든 스레드를 사용할 수 있습니다. 메모리 문제로 인해 간헐적인 충돌이 발생하는 경우 이 값을 줄이세요. - + Max Threads Export Profile entry in General 내보내기 시 사용할 최대 스레드 수 - + Text Detector Profile section title 글자 감지기 - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector 글자 감지 모델의 경로이며, 비워둘 때 내장 모델이 사용됩니다. 자동으로 다운로드 된 모델의 버전이 최신 버전이 아닐 때만 이 경로를 재정의하는 것이 좋습니다. 이 경로는 comictextdetector.pt (CUDA 사용)또는 comictextdetector.pt.onnx (CPU 사용)파일을 가리켜야합니다.<br>이전 버전의 모델 <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">다운로드</a> - + Model Path Profile entry in Text Detector 모델 경로 - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector 병렬 실행할 모델의 수입니다. 여러 모델을 동시에 실행할 수 있는 충분한 RAM(또는 CUDA의 경우에는 VRAM)이 있는 경우 유용합니다. 값을 늘리면 처리가 빨라질 수 있지만 하드웨어 사양에 비해 과하게 늘릴 때 충돌이 발생하거나 컴퓨터가 다운될 수 있습니다. 가용 메모리 2GB당, 또는 이미지 50개당 한 개의 모델을 쓰는 것이 좋습니다. 참고: 여러 모델을 시작하는 데 발생하는 오버헤드로 인해 50개 미만의 파일을 처리할 때는 이 설정은 무시됩니다.<br>이 설정은 전적으로 하드웨어의 성능에 의존하므로 <b>이에 대한 이슈를 열지 마세요</b>! - + Concurrent Models Profile entry in Text Detector 동시 실행 모델 개수 - + Preprocessor Profile section title 전처리기 - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor 글자 영역 크기는 총 픽셀 수로 지정됩니다. 따라서 200x200픽셀의 경우 200px * 200px = 40000px²입니다. - + Minimum size of any box to keep it. Profile comment in Preprocessor 지우지 않고 유지할 글자 영역의 최소 크기 값입니다. - + Box Min Size Profile entry in Preprocessor 글자 영역 최소 크기 - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor "unknown(알 수 없는)" 언어로 인식되는 글자 영역을 유지할 최소 크기. 알 수 없는 언어는 보통 로고나 잘못 인식된 텍스트 요소에서 발생합니다. - + Suspicious Box Min Size Profile entry in Preprocessor 알 수 없는 글자 영역 최소 크기 - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor 두 글자 영역을 하나로 합칠 최소한의 겹친 비율입니다. 이 비율(0-100)은 작은 영역이 큰 영역에 얼마나 겹쳐 포함되어 있는지에 대한 비율입니다. 값이 클수록 더 많이 겹쳐야 병합됩니다. - + Box Overlap Threshold Profile entry in Preprocessor 영역 겹침 기준값 - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor 지울 필요가 없는 글자 영역(숫자 혹은 특수문자만 있는 경우)을 OCR을 사용하여 감지할지 여부. - + OCR Enabled Profile entry in Preprocessor OCR 활성화됨 - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor 테서렉트를 사용하여 OCR을 수행할지 여부.<br>만약 활성화하면 사용할 수 있으면 테서렉트 OCR을 글자 감지에 사용합니다.<br>활성화하지 않으면 기본 내장 모델(manga-ocr)이 사용되며, 이는 세로쓰기로 된 일본어 텍스트에 가장 적합합니다. - + OCR Use Tesseract Profile entry in Preprocessor 테서렉트 OCR 사용 - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor OCR 작업 시 사용할 언어. 글자 감지기는 일본어와 영어만 감지할 수 있으며, 다른 모든 언어는 여기서 명시적으로 선택할 수 있습니다. 영역 단위 감지의 경우 감지기가 감지한 언어를 유지하며, 페이지 단위 감지의 경우 페이지 전역에서 가장 눈에 띄는 언어를 사용합니다.<br>참고: 기본적으로 일본어만 지원되며, 다른 모든 언어는 테서렉트를 활성화하고 관련 언어 팩을 설치해야합니다. - + OCR Language Profile entry in Preprocessor OCR 언어 - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor OCR 작업 시 사용할 엔진을 설정합니다.<br> -자동: 각 영역의 감지된 언어에 따라 자동으로 선택됩니다. 일본어는 manga-ocr을, 영어 또는 알 수 없는 텍스트에는 테서렉트를 사용합니다.<br> - manga ocr: 모든 텍스트 인식 작업에 대해 내장된 manga-ocr 모델을 사용하도록 강제합니다. 세로쓰기로 적힌 일본어 텍스트에 가장 적합합니다.<br> - 테서렉트: 모든 텍스트 인식 작업에 대해 테서렉트를 사용하도록 강제합니다. 영어 및 다른 언어들을 지원하는 범용 옵션입니다. - + OCR Engine Profile entry in Preprocessor OCR 엔진 - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor 페이지 전역에서 글자 영역을 처리하고 정렬하기 위한 읽는 순서를 지정합니다. 이는 개별 영역이 아닌 전체 페이지의 글자 영역 순서에 영향을 미칩니다. 이 전역 설정은 글자 영역이 어떤 순서로 정렬되고 후처리에 사용될지 결정합니다.<br> - 자동: 각 글자 영역에서 감지된 언어에 따라 자동으로 순서를 감지합니다.<br> - 망가: 오른쪽에서 왼쪽, 위에서 아래쪽 순서로, 일본 만화에 적합합니다.<br> - 코믹: 왼쪽에서 오른쪽, 위에서 아래쪽 순서로, 서양 만화에 적합합니다.<br>내용의 주요 배치 형식에 맞게 선택하세요. - + Reading Order Profile entry in Preprocessor 읽는 방향 - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor OCR을 수행할 글자 영역의 최대 크기입니다. OCR이 불필요한 영역은 보통 크고, OCR을 느리게 만들므로 이 설정으로 제한할 수 있습니다. - + OCR Max Size Profile entry in Preprocessor OCR 최대 크기 - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor OCR 결과에 비교할 정규식 패턴. 이 패턴에 일치하는 모든 인식 결과는 사용되지 않습니다.<br>참고: manga-ocr 모델은 결과로 전각 문자를 반환하므로 패턴도 이에 맞춰야합니다. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR 블랙리스트 패턴 - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor manga-ocr 모델은 일본어 텍스트만 처리할 수 있으므로 엄격 모드가 활성화되면 일본어임을 확신하지 못하는 글자 영역은 무시합니다. 간혹 숫자나 다른 특수 문자가 신뢰도를 떨어뜨려 언어를 감지할 수 없는 경우가 있는데, 엄격 모드를 비활성화하면 이러한 경우에는 무시되지 않습니다. 다른 언어라고 확신할 수 있는 영역들에 대해서는 이 설정과 관계없이 무시됩니다.<br>참고: 이 설정은 OCR 언어 옵션이 영역 단위 감지 혹은 페이지 단위 감지로 설정된 경우에만 유효합니다. - + OCR Strict Language Profile entry in Preprocessor OCR 언어 엄격 모드 - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor 글자 영역 주위에 추가할 여백입니다.이 여백은 글자 감지기 AI가 처음 만든 초기 글자 영역에 추가되며, 녹색으로 시각화됩니다. - + Box Padding Initial Profile entry in Preprocessor 초기 글자 영역 여백 - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor 영역의 오른쪽에 추가할 여백. 이 여백은 세로쓰기로 된 글자의 오른쪽에 있는 루비 텍스트를 덮는 데 도움이 됩니다. - + Box Right Padding Initial Profile entry in Preprocessor 초기 글자 영역 오른쪽 여백 - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor 글자 영역 주위에 추가할 여백입니다. 이 여백은 글자 감지기 AI의 오탐을 제거하는 데 사용되는 확장 글자 영역들에 추가되며, 보라색으로 시각화됩니다. - + Box Padding Extended Profile entry in Preprocessor 확장 글자 영역 여백 - + Box Right Padding Extended Profile entry in Preprocessor 확장 글자 영역 오른쪽 여백 - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor 글자 영역 주위에 추가할 여백입니다. 이 여백은 어떤 마스크를 사용할지 분석할 때 원본 이미지를 샘플링하는 데 사용되는 참조 영역에 추가되며, 파란색으로 시각화됩니다. - + Box Reference Padding Profile entry in Preprocessor 참조 영역 여백 - + Masker Profile section title 마스크 - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker 마스크 생성에 사용할 최대 스레드 수입니다. 이 값을 지정하지 않으면 사용할 수 있는 모든 스레드를 사용할 수 있습니다. 메모리 문제로 인해 간헐적인 충돌이 발생하는 경우 이 값을 줄이세요. - + Max Threads Profile entry in Masker 최대 스레드 수 - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker 단계마다 마스크를 몇 픽셀씩 늘릴지 정합니다. 이 값은 마스크의 주위를 늘리며 작은 값은 더 정확하지만 느립니다. - + Mask Growth Step Pixels Profile entry in Masker 단계별 확장 픽셀 - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker 마스크를 몇 단계까지 확장할지 정합니다. 값이 클수록 점점 더 큰 마스크를 만들 수 있지만 최종적으로 참조 영역 크기보다 커질 수 없습니다. - + Mask Growth Steps Profile entry in Masker 마스크 확장 단계 - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker 마스크의 최소 두께입니다. 이 값은 마스크의 첫 번째 단계 확장 시의 두께와 같으며 이후로는 단계별 확장 픽셀값을 따릅니다. 이렇게 하면 정확도를 위해 단계당 작은 확장 크기를 가질 수 있지만, 텍스트만 둘러싸는 매우 얇은 마스크를 방지할 수 있습니다. 이러한 경우에는 인페인팅이 더 적합한 선택일 수 있습니다. - + Min Mask Thickness Profile entry in Masker 최소 마스크 두께 - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter 마스크 주변 영역에 인페인팅을 수행하기 위해 주어진 마스크 가장자리 주변 색상의 최소 표준 편차입니다. 이 값이 마스크의 최대 편차와 일치하면 실패한 말풍선만 인페인팅되므로 다음 두 가지 설정은 무관합니다. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General 결과물을 레이어를 이용해 하나의 프로젝트 파일로 합칩니다. 현재 지원되는 파일 형식은 Photoshop PSD입니다.<br> - 없음: 각 이미지와 마스크는 기본 파일 형식으로 저장됩니다.<br> - 개별 PSD: 하나의 원본 이미지에 대한 이미지와 마스크를 하나의 PSD파일로 저장합니다.<br> - 단일 PSD: 각 원본 이미지에 대한 이미지와 마스크를 그룹화하여 최종적으로 하나의 PSD에 모든 이미지의 결과물을 저장합니다. - + Layered Export Profile entry in General 레이어 내보내기 - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker 컬러 마스크를 허용할지입니다. 허용하면 마스크에는 흰색, 회색, 검정색뿐만 아니라 모든 색상이 사용될 수 있습니다. - + Allow Colored Masks Profile entry in Masker 컬러 마스크 허용 - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker 픽셀을 미색으로 간주하기 위한 임곗값입니다. 마스크 가장자리의 중간 색상은 완벽한 흰색이 아닐 수 있으므로 완벽한 흰색 배경에 미색의 톤이 칠해지는 것을 방지하기 위해 이 값보다 밝은색은 모두 반올림되어 완벽한 흰색으로 처리됩니다. 값의 범위는 완전한 검정(0)에서 완전한 흰색(255)사이입니다. - + Off White Max Threshold Profile entry in Masker 미색 임곗값 - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker 마스크의 표준 편차는 마스크의 경계선에서 색상의 변화를 나타냅니다. 이를 위해 원본 이미지에서 마스크의 외부 경계에 있는 단일 픽셀 라인만 샘플링됩니다. 낮은 표준 편차는 마스크가 단색 위에 있음을 의미하며, 이는 마스크가 텍스트나 다른 객체와 겹쳐있지 않음을 나타냅니다. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker 마스크를 계속 확장하기 위한 표준 편차의 최소 개선 값입니다. 여기에서 더 높은 값을 설정하면, 더 큰 마스크를 고려하기 위해 더 큰 개선이 필요하게 되어, 작은 마스크를 우선시하는 효과를 줍니다. - + Mask Improvement Threshold Profile entry in Masker 마스크 개선 임곗값 - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker 빠른 마스크 선택 알고리즘을 사용할지입니다. 활성화하면 알고리즘은 첫 번째 완벽한 마스크를 선택합니다. 이 경우 속도는 빠르지만 조금 더 큰 마스크가 더 좋은 경우에는 가장 좋은 마스크를 찾지 못할 수도 있습니다. - + Mask Selection Fast Profile entry in Masker 빠른 마스크 선택 - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser 노이즈 감소에 사용할 최대 스레드 수입니다. 이 값을 지정하지 않으면 사용할 수 있는 모든 스레드를 사용할 수 있습니다. 메모리 문제로 인해 간헐적인 충돌이 발생하는 경우 이 값을 줄이세요. - + Max Threads Profile entry in Denoiser 최대 스레드 수 - + Inpainter Profile section title 인페인팅 - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter 실험적인 기능: 더 나은 기본 설정을 찾으시면 깃허브에 이슈를 올려 개선 사항을 모두와 공유해 주세요. 인페인팅은 기본적으로 활성화되어 있지 않습니다. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter 인페인팅은 머신 러닝을 사용하여 주변 영역을 기반으로 이미지의 영역을 대체하는 기술입니다. 제대로 지워지지 않았거나 아예 지워지지 않은 마스크의 경우 인페인팅을 사용할 수 있습니다. 노이즈 감소와 구분하기 위해 마스크가 여유 없이 덮고 있는 경우에 사용됩니다. 노이즈 감소가 된 마스크는 인페인팅이 되지 않습니다. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter 이 작업은 경우에 따라선 결과가 좋지 않을 수 있으므로 비활성화할 수 있습니다. 비활성화하려면 체크를 해제하세요. - + Inpainting Enabled Profile entry in Inpainter 인페인팅 활성화 - + Inpainting Min Std Dev Profile entry in Inpainter 인페인팅 최소 표준 편차 - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter 인페인팅을 수행할 마스크의 최대 반경입니다. 이보다 큰 마스크에 대해서는 처리하지 않습니다. 여백이 이 값보다 크면 이미 충분히 좋은 상태일 가능성이 높기 때문입니다. - + Inpainting Max Mask Radius Profile entry in Inpainter 인페인팅 최대 마스크 반경 - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter 인페인팅할 마스크 주변 최소 반경입니다. 이 값은 최적의 마스크 크기에 추가되어 인페인팅이 마스크 전체를 덮도록 합니다. - + Min Inpainting Radius Profile entry in Inpainter 최소 인페인팅 반경 - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter 지우기 어려운 것으로 판정되는 마스크, 즉 표준 편차가 높은 마스크는 인페인팅 반경을 늘려 마스크가 더 많은 부분을 덮을 수 있도록 합니다. 이것은 최소 인페인팅 반경에 추가 여백을 더하는 것으로, 인페인팅 반경 배수 * 마스크 표준 편차로 계산됩니다 - + Inpainting Radius Multiplier Profile entry in Inpainter 인페인팅 반경 배수 - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter 인페인팅할 마스크 주변 최대 반경입니다. 큰 반경 배수를 사용할 경우 이미지를 과하게 덮는 것을 방지하기 위해 인페인팅 영역이 최대로 덮을 수 있는 크기를 제한합니다. - + Max Inpainting Radius Profile entry in Inpainter 최대 인페인팅 반경 - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter 인페인팅 이후 나머지 이미지에 영향을 미치지 않도록 원본 이미지에서 결과물을 잘라냅니다. 이렇게 하면 원본 이미지가 최대한 보존됩니다. 이 반경은 인페인팅 모델이 전용 영역 외부의 몇 픽셀을 수정하기 때문에 최종 인페인팅 반경 주위에 추가됩니다. - + Inpainting Isolation Radius Profile entry in Inpainter 인페인팅 격리 반경 - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter 인페인팅 된 마스크의 가장자리를 부드럽게 만들어 인페인팅된 부분과 원본 이미지를 자연스럽게 혼합합니다.<br>인페인팅 작업 후에 윤곽선이 보이면 최소 인페인팅 값을 이 값의 절반으로 설정하세요. - + Inpainting Fade Radius Profile entry in Inpainter 인페인팅 혼합 반경 - + Detect per box Language option for OCR 영역 단위 감지 - + Detect per page Language option for OCR 페이지 단위 감지 - + Japanese Language option for OCR 일본어 - + English Language option for OCR 영어 - + Korean Language option for OCR 한국어 - + Korean (vertical) Language option for OCR 한국어(세로쓰기) - + Chinese - Simplified Language option for OCR 중국어(간체) - + Chinese - Traditional Language option for OCR 중국어(번체) - + Albanian Language option for OCR 알바니아어 - + Arabic Language option for OCR 아랍어 - + Azerbaijani Language option for OCR 아제르바이잔어 - + Azerbaijani - Cyrilic Language option for OCR 아제르바이잔어(키릴문자) - + Bengali Language option for OCR 벵골어 - + Bulgarian Language option for OCR 불가리아어 - + Burmese Language option for OCR 버마어 - + Catalan; Valencian Language option for OCR 카탈루냐어(발렌시아어) - + Croatian Language option for OCR 크로아티아어 - + Czech Language option for OCR 체코어 - + Danish Language option for OCR 덴마크어 - + Dutch; Flemish Language option for OCR 네덜란드어(플라망어) - + Esperanto Language option for OCR 에스페란토어 - + Estonian Language option for OCR 에스토니아어 - + Finnish Filipino Language option for OCR 핀란드어 - + French Language option for OCR 프랑스어 - + Georgian Language option for OCR 조지아어 - + German Language option for OCR 독일어 - + Greek Language option for OCR 그리스어 - + Hebrew Language option for OCR 히브리어 - + Hindi Language option for OCR 힌디어 - + Hungarian Language option for OCR 헝가리어 - + Indonesian Language option for OCR 인도네시아어 - + Italian Language option for OCR 이탈리아어 - + Kazakh Language option for OCR 카자흐어 - + Latin Language option for OCR 라틴어 - + Lithuanian Language option for OCR 리투아니아어 - + Luxembourgish Language option for OCR 룩셈부르크어 - + Malay Language option for OCR 말레이어 - + Mongolian Language option for OCR 몽골어 - + Nepali Language option for OCR 네팔어 - + Norwegian Language option for OCR 노르웨이어 - + Persian Language option for OCR 페르시아어 - + Polish Language option for OCR 폴란드어 - + Portuguese Language option for OCR 포르투갈어 - + Romanian; Moldavian Language option for OCR 루마니아어(몰도바어) - + Russian Language option for OCR 러시아어 - + Serbian Language option for OCR 세르비아어 - + Serbian - Latin Language option for OCR 세르비아어(라틴문자) - + Slovak Language option for OCR 슬로바키아어 - + Slovenian Language option for OCR 슬로베니아어 - + Spanish; Castilian Language option for OCR 스페인어(카스티야어) - + Swedish Language option for OCR 스웨덴어 - + Tagalog Language option for OCR 타갈로그어 - + Tamil Language option for OCR 타밀어 - + Telugu Language option for OCR 텔루구어 - + Thai Language option for OCR 태국어 - + Turkish Language option for OCR 터키어 - + Ukrainian Language option for OCR 우크라이나어 - + Vietnamese Language option for OCR 베트남어 - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker 고려할 마스크의 최대 표준 편차입니다. 여기서 값이 높을수록 텍스트나 다른 물체와 교차하는 마스크의 허용 오차가 커지며, 이는 어차피 다시 칠해야 하므로 좋은 마스크가 아닙니다. 이 값을 0으로 설정하면 완벽한 마스크만 허용되며, 이는 매우 높은 해상도의 이미지에서 권장됩니다. - + Mask Max Standard Deviation Profile entry in Masker 마스크 최대 표준 편차 - + Color to use for the debug mask. Profile comment in Masker 디버깅에 사용할 컬러 마스크의 색상. - + Debug Mask Color Profile entry in Masker 디버그 컬러 마스크 - + Denoiser Profile section title 노이즈 감소 - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser 글자를 마스킹하기 어려운 경우 작은 마스크가 필요할 수 있으며, 이에 따라 글자 주변에 많은 JPEG 아티팩트(노이즈)가 남을 수 있습니다. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser 이는 마스크 주변의 픽셀 색상이 균일하지 않다는 것을 의미하며, 표준 편차로 정량화됩니다. 노이즈 감소는 표준 편차가 최소인 마스크를 대상으로서 마스크 주변 영역의 노이즈를 제거할 수 있습니다. 이렇게 하면 이미지의 나머지 부분의 디테일은 유지되면서 텍스트가 있던 바로 주변의 아티팩트는 제거됩니다. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser 이 작업은 선택 사항이며 JPEG 아티팩트가 없는 고해상도 이미지의 경우 불필요할 수도 있으므로 여기에서 비활성화할 수 있습니다. 노이즈 감소를 비활성화하려면 체크를 해제합니다. - + Denoising Enabled Profile entry in Denoiser 노이즈 감소 활성화 - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser 마스크 주변 영역에서 노이즈 감소를 수행하기 위해 지정된 마스크 가장자리 주변 색상의 최소 표준 편차입니다. - + Noise Min Standard Deviation Profile entry in Denoiser 노이즈 최소 표준 편차 - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser 참고: 인페인팅이 활성화된 경우 인페인팅 최소 표준 편차가 이 마스크의 최댓값으로 작용하며, 그 이후에는 이 마스크가 인페인터에게 전달됩니다. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser 마스크 주변의 노이즈 감소를 수행할 범위의 두께입니다. - + Noise Outline Size Profile entry in Denoiser 노이즈 감소 범위 - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser 노이즈 감소 마스크의 가장자리를 값만큼 부드럽게 하여 노이즈가 제거된 부분을 이미지의 원래 영역과 자연스럽게 합칩니다. - + Noise Fade Radius Profile entry in Denoiser 혼합 반경 - + Set to True to support colored images. Profile comment in Denoiser 컬러 이미지를 지원하려면 활성화합니다. - + Colored Images Profile entry in Denoiser 컬러 이미지 지원 - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser 노이즈 감소 강도를 설정합니다. 값이 클수록 노이즈가 더 많이 제거됩니다. - + Filter Strength Profile entry in Denoiser 필터 강도 - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser 밝기에 대비하여 색상 채널의 노이즈를 얼마나 강하게 제거할지 설정합니다. 값이 높을수록 노이즈가 더 많이 제거됩니다. - + Color Filter Strength Profile entry in Denoiser 컬러 필터 강도 - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser 가중치를 계산하는 데 사용되는 컨텍스트 영역의 픽셀 단위 크기입니다. 홀수여야 합니다. - + Template Window Size Profile entry in Denoiser 컨텍스트 영역 크기 - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser 특정 픽셀에 대한 가중 평균을 계산하는 데 사용되는 픽셀 단위의 영역 크기입니다. 홀수여야 합니다. - + Search Window Size Profile entry in Denoiser 검색 영역 크기 @@ -4242,7 +4247,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. 모든 스레드 @@ -4659,14 +4664,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>글자 감지 모델의 경로 {value}가 존재하지 않아 기본값으로 변경됩니다. 모델은 <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">이곳</a>에서 직접 다운로드하거나, 기본 모델을 이용하여 진행할 수 있습니다.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. {key}를 위한 정규식 패턴 "{value}"은 유효하지 않아 기본값으로 변경됩니다. From 4f01ae15c23f046586cac59b99fbca31b0f38c4b Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:19 +0100 Subject: [PATCH 13/18] New translations panelcleaner_source.ts (Portuguese) --- translations/PanelCleaner_pt_PT.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_pt_PT.ts b/translations/PanelCleaner_pt_PT.ts index ff3df4d..0aec83d 100644 --- a/translations/PanelCleaner_pt_PT.ts +++ b/translations/PanelCleaner_pt_PT.ts @@ -4,7 +4,7 @@ - + default reserved profile name default @@ -20,54 +20,54 @@ Profile {profile_name} created. - + Profile name cannot be empty. Profile name cannot be empty. - + Profile name already in use. Profile name already in use. - + Profile name is reserved. Profile name is reserved. - - - - + + + + Unknown Unknown - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Grayscale - + 1-bit Color mode 1-bit - + Palette Color mode Palette @@ -78,45 +78,45 @@ Error downloading file from url: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Error downloading file from url: {url} Failed to save the file to {save_path} - + Error downloading file from url: {url} The file content is different from expected. Error downloading file from url: {url} The file content is different from expected. + - - - + + + - - - + + N/A N/A - + OCR Analytics OCR Analytics - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) @@ -155,66 +155,66 @@ Removed bubbles: - + Small boxes Small boxes - + Removed boxes Removed boxes - - + + Failed Failed - + Perfect Perfect - + Total Total - + Mask Fitment Analytics Mask Fitment Analytics - + Total Total - + Total boxes Total boxes - + Masks succeeded Masks succeeded - + Masks failed Masks failed - + Perfect masks Perfect masks - + Average border deviation Average border deviation @@ -233,7 +233,7 @@ Mask usage by thickness (in pixels): Mask ({thickness}px) - + Box mask Box mask @@ -254,27 +254,27 @@ Pages with failures / total: - + Denoising Analytics Denoising Analytics - + Total masks Total masks - + Masks denoised Masks denoised - + Minimum deviation to denoise Minimum deviation to denoise - + Maximum allowed deviation Maximum allowed deviation @@ -293,27 +293,27 @@ Pages with failures / total: Denoised - + Inpainting Analytics Inpainting Analytics - + Inpainting performed Inpainting performed - + Average thickness Average thickness - + Minimum thickness Minimum thickness - + Maximum thickness Maximum thickness @@ -372,7 +372,7 @@ Pages with failures / total: The file format was not recognized. - + The CSV file must start with a header row, followed by data rows. The CSV file must start with a header row, followed by data rows. @@ -402,7 +402,7 @@ Pages with failures / total: Multiple images match the file path: {path} - + Error, no version found. Error, no version found. @@ -412,7 +412,7 @@ Pages with failures / total: CUDA Error - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pages with failures / total: Inpainting model not found. - + filename filename - + startx startx - + starty starty - + endx endx - + endy endy - + text text - + Saved detected text to {output_file} Saved detected text to {output_file} - + Failed to write detected text to {output_file} Failed to write detected text to {output_file} - + Save Failed Save Failed - + Failed to write detected text to file. Failed to write detected text to file. @@ -538,7 +538,7 @@ Pages with failures / total: - + Page layered export @@ -625,7 +625,7 @@ An AI-powered tool to clean manga panels. ColorButton - + Select Color Select Color @@ -671,7 +671,7 @@ An AI-powered tool to clean manga panels. Unsupported system - + This system is not supported. This system is not supported. @@ -686,12 +686,12 @@ An AI-powered tool to clean manga panels. Uninstall for Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. The extension will be installed at {path}. This doesn't require root permissions. - + The extension file will be deleted. This doesn't require root permissions. The extension file will be deleted. This doesn't require root permissions. @@ -706,17 +706,17 @@ An AI-powered tool to clean manga panels. Uninstall for Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. The extension will be installed in the Windows registry.This requires admin permissions. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. The extension will be uninstalled from the Windows registry.This requires admin permissions. - + Dolphin extension installed successfully. Dolphin extension installed successfully. @@ -726,7 +726,7 @@ An AI-powered tool to clean manga panels. Installation successful - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Dolphin extension installed successfully. @@ -806,12 +806,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) Original size in pixels (width × height) - + Processing size in pixels (width × height), scale factor Processing size in pixels (width × height), scale factor @@ -837,7 +837,7 @@ By default, this uses the built-in profile. However, you can change this default Unsupported TIFF files - + The following 5-channel TIFF files are not supported: The following 5-channel TIFF files are not supported: @@ -864,27 +864,27 @@ By default, this uses the built-in profile. However, you can change this default split - + Number of boxes removed by the OCR model / total boxes Number of boxes removed by the OCR model / total boxes - + Number of boxes that failed to generate a mask / total boxes Number of boxes that failed to generate a mask / total boxes - + Number of boxes that were perfectly masked / total boxes Number of boxes that were perfectly masked / total boxes - + Number of boxes that were denoised / total boxes Number of boxes that were denoised / total boxes - + Number of boxes that were inpainted Number of boxes that were inpainted @@ -1032,14 +1032,14 @@ By default, this uses the built-in profile. However, you can change this default Impossible Request - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting is disabled in the current profile, this output can't be generated. @@ -1061,7 +1061,7 @@ Please enable inpainting in the profile settings and try again. Export failed - + Failed to export image: Failed to export image: @@ -1104,7 +1104,7 @@ Please enable inpainting in the profile settings and try again. result - + results results @@ -1119,17 +1119,17 @@ Please enable inpainting in the profile settings and try again. new image - + new images new images - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. @@ -1228,12 +1228,12 @@ Please enable inpainting in the profile settings and try again. Note: Name "{name}" was hidden - + Corrupted log session Corrupted log session - + Current session Current session @@ -1248,7 +1248,7 @@ Please enable inpainting in the profile settings and try again. Error - + Errors Errors @@ -1258,7 +1258,7 @@ Please enable inpainting in the profile settings and try again. Critical - + Criticals Criticals @@ -1274,49 +1274,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button Open Config - + Open Log Statusbar button Open Log - + CUDA Enabled CUDA Enabled - + Select Output Directory Select Output Directory - + Aborting... Aborting... - + Loading OCR model... Loading OCR model... - + Loaded OCR model. Loaded OCR model. - + Error Error - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1325,7 +1325,7 @@ Continue anyway? Continue anyway? - + Delete Models Delete Models @@ -1335,23 +1335,23 @@ Continue anyway? Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. - + Encountered error: Encountered error: - + MPS Enabled MPS Enabled - - + + Inpainting Model Update Inpainting Model Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1365,148 +1365,148 @@ Switch to the new model? Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. - + Layout Reset Layout Reset - + The window layout has been reset. The default layout will be used next time. The window layout has been reset. The default layout will be used next time. - + Multiple Instances Multiple Instances - + Found process ID in lock file: Found process ID in lock file: - + Failed to delete models. Failed to delete models. - + Delete Failed Delete Failed - + Models Deleted Models Deleted - + The models were deleted. Would you like to download them again? The models were deleted. Would you like to download them again? - + Not Supported Not Supported - + Unsupported system or package format Unsupported system or package format - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Restart Required - + The language has been changed. Please restart the application for the changes to take effect. The language has been changed. Please restart the application for the changes to take effect. - - + + System Language System Language - + Import Profile Import Profile - + Profile Files (*.conf) Profile Files (*.conf) - + Profile Imported Profile Imported - + Import Error Import Error - + Failed to Delete Failed to Delete - + The default profile cannot be deleted. The default profile cannot be deleted. - + Delete Profile Delete Profile - + Are you sure you want to delete the profile {profile_name}? Are you sure you want to delete the profile {profile_name}? - + Delete Error Delete Error - + Failed to delete the profile. Failed to delete the profile. @@ -1518,217 +1518,217 @@ Switching profiles will discard changes to the current profile. Switching profiles will discard changes to the current profile. - + Unsaved changes Unsaved changes - - + + Load Error Load Error - - + + Failed to load profile: Failed to load profile: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Profile Created - + Create Error Create Error - + Failed to save profile to {profile_path} Failed to save profile to {profile_path} - - + + Save Error Save Error - + Failed to save profile. Failed to save profile. - + Profile saved to {profile_path} Profile saved to {profile_path} - + Failed to save config. Failed to save config. - + Failed to save the new profile to the configuration file. Failed to save the new profile to the configuration file. - - + + No Files No Files - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. - + No Outputs No Outputs - + No outputs were requested. Please select at least one output before cleaning. No outputs were requested. Please select at least one output before cleaning. - - + + cleaned cleaned - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists File Exists - + The file '{output_path}' already exists. Overwrite? The file '{output_path}' already exists. Overwrite? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Processing Finished - + Finished processing all files. Finished processing all files. - + Processing Aborted Processing Aborted - + Processing aborted. Processing aborted. - + Processing Error Processing Error - + Encountered an error while processing files. Encountered an error while processing files. @@ -1739,7 +1739,7 @@ Check the online documentation for how to add support for more languages. - + Profile Profile @@ -1754,360 +1754,365 @@ Check the online documentation for how to add support for more languages.Apply - - + + Save Save - + Reset All Reset All - + Drag and Drop Images or Folders Here Drag and Drop Images or Folders Here - + Images Images - - + + File File - + Size Size - + Processing Size The size that the picture is shrunk down to for processing. Processing Size - + File Size File Size - + Color Mode Like RGB, grayscale etc. Color Mode - + Analytics Can also call this statistics. Analytics - + Process Process - + Cleaning Cleaning - + OCR OCR - + Edit Existing Output Edit Existing Output - + Abort Abort - + Start Start - + Output Output - + Cleaned Image Cleaned Image - + Mask Mask - + Isolated Text Isolated Text - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Save Output - + Output Directory: Output Directory: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. You can use a relative path to create a subfolder at the image's original location, or use an absolute path. - + Plain Text Plain Text - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Output File: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Generating: - + Current Step: Current Step: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Images Processed: - + Set Default Set Default - + Settings Settings - + Theme As in color theme Theme - + Language Language - + Help Help - + Add Files... Add Files... - + Remove All Files Remove All Files - + Remove File Remove File - + Help Translate Panel Cleaner Help Translate Panel Cleaner - + Report an Issue... Report an Issue... - + Simulate Exception Simulate Exception - + Add File Browser Extension... Add File Browser Extension... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New New - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Delete - + Import... Import... - + Online Documentation Online Documentation - + About Panel Cleaner About Panel Cleaner - + View License View License - + Save as... Save as... - + System System - + Dark Dark - + Light Light - + Add Folder... Add Folder... - + Delete Machine Learning Models Delete Machine Learning Models - + Download Machine Learning Models Download Machine Learning Models - + Donate Donate @@ -2139,19 +2144,19 @@ Check the online documentation for how to add support for more languages.OCR model: - + Already downloaded Already downloaded - + Text Detector model (CUDA) Text Detector model (CUDA) - + Text Detector model (CPU) Text Detector model (CPU) @@ -2170,7 +2175,7 @@ Check the online documentation for how to add support for more languages.OCR download errors - + ETA estimated time of completion ETA @@ -2748,7 +2753,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2829,73 +2834,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input The original image with scaling applied (if needed). - + Input Step name in the image details view Input - + Input Output name in the image details view Input - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask The rough mask generated by the AI. - + Text Detection Step name in the image details view Text Detection - + AI Mask Output name in the image details view AI Mask - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes The unfiltered box data generated by the AI. - + Raw Boxes Output name in the image details view Raw Boxes - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes The outlines of the text boxes the AI found. - + Preprocessor Step name in the image details view Preprocessor - + Initial Boxes Output name in the image details view Initial Boxes - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2903,181 +2908,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. - + Final Boxes Output name in the image details view Final Boxes - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask The mask of the merged boxes. - + Masker Step name in the image details view Masker - + Box Mask Output name in the image details view Box Mask - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask The rough text detection mask with everything outside the box mask cut out. - + Cut Mask Output name in the image details view Cut Mask - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers The different steps of growth around the cut mask displayed in different colors. - + Mask Layers Output name in the image details view Mask Layers - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). - + Fitment Quality Output name in the image details view Fitment Quality - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay The input image with the final mask overlaid in color. - + Mask Overlay Output name in the image details view Mask Overlay - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask The collection of masks for each bubble that fit best. - + Final Mask Output name in the image details view Final Mask - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text The text layer isolated from the input image. - + Isolated Text Output name in the image details view Isolated Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output The input image with the final mask applied. - + Masked Output Output name in the image details view Masked Output - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask The masks that required denoising, to be overlaid on the final mask when exporting. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Denoise Mask - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output The input image with the denoised mask applied. - + Denoised Output Output name in the image details view Denoised Output - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask The inpainted sections around the text that was poorly cleaned, if at all. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Inpainted Mask - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output The input image: cleaned, denoised (if enabled), and inpainted. - + Inpainted Output Output name in the image details view Inpainted Output - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3086,1141 +3091,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title General - + About this profile: Profile comment in General About this profile: - + Notes Profile entry in General Notes - + Preferred file type to save the cleaned image as. Profile comment in General Preferred file type to save the cleaned image as. - + Preferred File Type Profile entry in General Preferred File Type - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Preferred file type to save the mask as. Only image formats that allow for transparency are supported. - + Preferred Mask File Type Profile entry in General Preferred Mask File Type - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. - + Input Height Lower Target Profile entry in General Input Height Lower Target - + Input Height Upper Target Profile entry in General Input Height Upper Target - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Text Detector - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> - + Model Path Profile entry in Text Detector Model Path - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! - + Concurrent Models Profile entry in Text Detector Concurrent Models - + Preprocessor Profile section title Preprocessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Minimum size of any box to keep it. - + Box Min Size Profile entry in Preprocessor Box Min Size - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. - + Suspicious Box Min Size Profile entry in Preprocessor Suspicious Box Min Size - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. - + Box Overlap Threshold Profile entry in Preprocessor Box Overlap Threshold - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. - + OCR Enabled Profile entry in Preprocessor OCR Enabled - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. - + OCR Use Tesseract Profile entry in Preprocessor OCR Use Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. - + OCR Engine Profile entry in Preprocessor OCR Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. - + Reading Order Profile entry in Preprocessor Reading Order - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. - + OCR Max Size Profile entry in Preprocessor OCR Max Size - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR Blacklist Pattern - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor OCR Strict Language - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. - + Box Padding Initial Profile entry in Preprocessor Box Padding Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. - + Box Right Padding Initial Profile entry in Preprocessor Box Right Padding Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Box Right Padding Extended - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. - + Box Reference Padding Profile entry in Preprocessor Box Reference Padding - + Masker Profile section title Masker - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. - + Mask Growth Step Pixels Profile entry in Masker Mask Growth Step Pixels - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. - + Mask Growth Steps Profile entry in Masker Mask Growth Steps - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. - + Min Mask Thickness Profile entry in Masker Min Mask Thickness - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Off White Max Threshold - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. - + Mask Improvement Threshold Profile entry in Masker Mask Improvement Threshold - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. - + Mask Selection Fast Profile entry in Masker Mask Selection Fast - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting Enabled - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. - + Min Inpainting Radius Profile entry in Inpainter Min Inpainting Radius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation - + Inpainting Radius Multiplier Profile entry in Inpainter Inpainting Radius Multiplier - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. - + Max Inpainting Radius Profile entry in Inpainter Max Inpainting Radius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. - + Inpainting Isolation Radius Profile entry in Inpainter Inpainting Isolation Radius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. - + Mask Max Standard Deviation Profile entry in Masker Mask Max Standard Deviation - + Color to use for the debug mask. Profile comment in Masker Color to use for the debug mask. - + Debug Mask Color Profile entry in Masker Debug Mask Color - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. - + Denoising Enabled Profile entry in Denoiser Denoising Enabled - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. - + Noise Min Standard Deviation Profile entry in Denoiser Noise Min Standard Deviation - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser The thickness of an outline to denoise around a mask. - + Noise Outline Size Profile entry in Denoiser Noise Outline Size - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. - + Noise Fade Radius Profile entry in Denoiser Noise Fade Radius - + Set to True to support colored images. Profile comment in Denoiser Set to True to support colored images. - + Colored Images Profile entry in Denoiser Colored Images - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the image. Higher values mean more denoising. - + Filter Strength Profile entry in Denoiser Filter Strength - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. - + Color Filter Strength Profile entry in Denoiser Color Filter Strength - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Size in pixels of the template patch that is used to compute weights. Should be odd. - + Template Window Size Profile entry in Denoiser Template Window Size - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. - + Search Window Size Profile entry in Denoiser Search Window Size @@ -4247,7 +4252,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4664,14 +4669,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From cdb9d8611edc3df483699e8fedd4befee011247c Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:20 +0100 Subject: [PATCH 14/18] New translations panelcleaner_source.ts (Turkish) --- translations/PanelCleaner_tr_TR.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_tr_TR.ts b/translations/PanelCleaner_tr_TR.ts index f5804d4..86e1201 100644 --- a/translations/PanelCleaner_tr_TR.ts +++ b/translations/PanelCleaner_tr_TR.ts @@ -4,7 +4,7 @@ - + default reserved profile name default @@ -20,54 +20,54 @@ Profile {profile_name} created. - + Profile name cannot be empty. Profile name cannot be empty. - + Profile name already in use. Profile name already in use. - + Profile name is reserved. Profile name is reserved. - - - - + + + + Unknown Unknown - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Grayscale - + 1-bit Color mode 1-bit - + Palette Color mode Palette @@ -78,45 +78,45 @@ Error downloading file from url: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Error downloading file from url: {url} Failed to save the file to {save_path} - + Error downloading file from url: {url} The file content is different from expected. Error downloading file from url: {url} The file content is different from expected. + - - - + + + - - - + + N/A N/A - + OCR Analytics OCR Analytics - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) @@ -155,66 +155,66 @@ Removed bubbles: - + Small boxes Small boxes - + Removed boxes Removed boxes - - + + Failed Failed - + Perfect Perfect - + Total Total - + Mask Fitment Analytics Mask Fitment Analytics - + Total Total - + Total boxes Total boxes - + Masks succeeded Masks succeeded - + Masks failed Masks failed - + Perfect masks Perfect masks - + Average border deviation Average border deviation @@ -233,7 +233,7 @@ Mask usage by thickness (in pixels): Mask ({thickness}px) - + Box mask Box mask @@ -254,27 +254,27 @@ Pages with failures / total: - + Denoising Analytics Denoising Analytics - + Total masks Total masks - + Masks denoised Masks denoised - + Minimum deviation to denoise Minimum deviation to denoise - + Maximum allowed deviation Maximum allowed deviation @@ -293,27 +293,27 @@ Pages with failures / total: Denoised - + Inpainting Analytics Inpainting Analytics - + Inpainting performed Inpainting performed - + Average thickness Average thickness - + Minimum thickness Minimum thickness - + Maximum thickness Maximum thickness @@ -372,7 +372,7 @@ Pages with failures / total: The file format was not recognized. - + The CSV file must start with a header row, followed by data rows. The CSV file must start with a header row, followed by data rows. @@ -402,7 +402,7 @@ Pages with failures / total: Multiple images match the file path: {path} - + Error, no version found. Error, no version found. @@ -412,7 +412,7 @@ Pages with failures / total: CUDA Error - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pages with failures / total: Inpainting model not found. - + filename filename - + startx startx - + starty starty - + endx endx - + endy endy - + text text - + Saved detected text to {output_file} Saved detected text to {output_file} - + Failed to write detected text to {output_file} Failed to write detected text to {output_file} - + Save Failed Save Failed - + Failed to write detected text to file. Failed to write detected text to file. @@ -538,7 +538,7 @@ Pages with failures / total: - + Page layered export @@ -625,7 +625,7 @@ Manga panellerini temizlemek için yapay zeka destekli bir araç. ColorButton - + Select Color Select Color @@ -671,7 +671,7 @@ Manga panellerini temizlemek için yapay zeka destekli bir araç. Unsupported system - + This system is not supported. This system is not supported. @@ -686,12 +686,12 @@ Manga panellerini temizlemek için yapay zeka destekli bir araç. Uninstall for Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. The extension will be installed at {path}. This doesn't require root permissions. - + The extension file will be deleted. This doesn't require root permissions. The extension file will be deleted. This doesn't require root permissions. @@ -706,17 +706,17 @@ Manga panellerini temizlemek için yapay zeka destekli bir araç. Uninstall for Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. The extension will be installed in the Windows registry.This requires admin permissions. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. The extension will be uninstalled from the Windows registry.This requires admin permissions. - + Dolphin extension installed successfully. Dolphin extension installed successfully. @@ -726,7 +726,7 @@ Manga panellerini temizlemek için yapay zeka destekli bir araç. Installation successful - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Dolphin extension installed successfully. @@ -806,12 +806,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) Original size in pixels (width × height) - + Processing size in pixels (width × height), scale factor Processing size in pixels (width × height), scale factor @@ -837,7 +837,7 @@ By default, this uses the built-in profile. However, you can change this default Unsupported TIFF files - + The following 5-channel TIFF files are not supported: The following 5-channel TIFF files are not supported: @@ -864,27 +864,27 @@ By default, this uses the built-in profile. However, you can change this default split - + Number of boxes removed by the OCR model / total boxes Number of boxes removed by the OCR model / total boxes - + Number of boxes that failed to generate a mask / total boxes Number of boxes that failed to generate a mask / total boxes - + Number of boxes that were perfectly masked / total boxes Number of boxes that were perfectly masked / total boxes - + Number of boxes that were denoised / total boxes Number of boxes that were denoised / total boxes - + Number of boxes that were inpainted Number of boxes that were inpainted @@ -1032,14 +1032,14 @@ By default, this uses the built-in profile. However, you can change this default Impossible Request - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting is disabled in the current profile, this output can't be generated. @@ -1061,7 +1061,7 @@ Please enable inpainting in the profile settings and try again. Export failed - + Failed to export image: Failed to export image: @@ -1104,7 +1104,7 @@ Please enable inpainting in the profile settings and try again. result - + results results @@ -1119,17 +1119,17 @@ Please enable inpainting in the profile settings and try again. new image - + new images new images - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. @@ -1228,12 +1228,12 @@ Please enable inpainting in the profile settings and try again. Note: Name "{name}" was hidden - + Corrupted log session Corrupted log session - + Current session Current session @@ -1248,7 +1248,7 @@ Please enable inpainting in the profile settings and try again. Error - + Errors Errors @@ -1258,7 +1258,7 @@ Please enable inpainting in the profile settings and try again. Critical - + Criticals Criticals @@ -1274,49 +1274,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button Open Config - + Open Log Statusbar button Open Log - + CUDA Enabled CUDA Enabled - + Select Output Directory Select Output Directory - + Aborting... Aborting... - + Loading OCR model... Loading OCR model... - + Loaded OCR model. Loaded OCR model. - + Error Error - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1325,7 +1325,7 @@ Continue anyway? Continue anyway? - + Delete Models Delete Models @@ -1335,23 +1335,23 @@ Continue anyway? Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. - + Encountered error: Encountered error: - + MPS Enabled MPS Enabled - - + + Inpainting Model Update Inpainting Model Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1365,148 +1365,148 @@ Switch to the new model? Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. - + Layout Reset Layout Reset - + The window layout has been reset. The default layout will be used next time. The window layout has been reset. The default layout will be used next time. - + Multiple Instances Multiple Instances - + Found process ID in lock file: Found process ID in lock file: - + Failed to delete models. Failed to delete models. - + Delete Failed Delete Failed - + Models Deleted Models Deleted - + The models were deleted. Would you like to download them again? The models were deleted. Would you like to download them again? - + Not Supported Not Supported - + Unsupported system or package format Unsupported system or package format - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Restart Required - + The language has been changed. Please restart the application for the changes to take effect. The language has been changed. Please restart the application for the changes to take effect. - - + + System Language System Language - + Import Profile Import Profile - + Profile Files (*.conf) Profile Files (*.conf) - + Profile Imported Profile Imported - + Import Error Import Error - + Failed to Delete Failed to Delete - + The default profile cannot be deleted. The default profile cannot be deleted. - + Delete Profile Delete Profile - + Are you sure you want to delete the profile {profile_name}? Are you sure you want to delete the profile {profile_name}? - + Delete Error Delete Error - + Failed to delete the profile. Failed to delete the profile. @@ -1518,217 +1518,217 @@ Switching profiles will discard changes to the current profile. Switching profiles will discard changes to the current profile. - + Unsaved changes Unsaved changes - - + + Load Error Load Error - - + + Failed to load profile: Failed to load profile: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Profile Created - + Create Error Create Error - + Failed to save profile to {profile_path} Failed to save profile to {profile_path} - - + + Save Error Save Error - + Failed to save profile. Failed to save profile. - + Profile saved to {profile_path} Profile saved to {profile_path} - + Failed to save config. Failed to save config. - + Failed to save the new profile to the configuration file. Failed to save the new profile to the configuration file. - - + + No Files No Files - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. - + No Outputs No Outputs - + No outputs were requested. Please select at least one output before cleaning. No outputs were requested. Please select at least one output before cleaning. - - + + cleaned temizlenmiş - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists File Exists - + The file '{output_path}' already exists. Overwrite? The file '{output_path}' already exists. Overwrite? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Processing Finished - + Finished processing all files. Finished processing all files. - + Processing Aborted Processing Aborted - + Processing aborted. Processing aborted. - + Processing Error Processing Error - + Encountered an error while processing files. Encountered an error while processing files. @@ -1739,7 +1739,7 @@ Check the online documentation for how to add support for more languages. - + Profile Profil @@ -1754,360 +1754,365 @@ Check the online documentation for how to add support for more languages.Uygula - - + + Save Kaydet - + Reset All Tümünü Sıfırla - + Drag and Drop Images or Folders Here Görüntüleri veya Klasörleri Buraya Sürükleyip Bırakın - + Images Görüntüler - - + + File Dosya - + Size Boyut - + Processing Size The size that the picture is shrunk down to for processing. İşlenme Boyutu - + File Size Dosya Boyutu - + Color Mode Like RGB, grayscale etc. Renk Modu - + Analytics Can also call this statistics. Analitikler - + Process İşlem - + Cleaning Temizleme - + OCR OCR - + Edit Existing Output Edit Existing Output - + Abort İptal et - + Start Başlat - + Output Çıktı - + Cleaned Image Temizlenmiş Görüntü - + Mask Maske - + Isolated Text İzole Metin - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Çıktıyı Kaydet - + Output Directory: Çıktı Dizini: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. Görüntüsünün orijinal konumunda bir alt klasör oluşturmak için göreli bir yol kullanabilir veya mutlak bir yol kullanabilirsiniz. - + Plain Text Düz Metin - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Çıktı Dosyası: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Bir dosya adıyla birlikte mutlak veya göreli bir yol girin. Tüm görüntülerin çıktısı aynı dosyaya yazılır.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Üretiliyor: - + Current Step: Şu anki adım: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. İşlenen görüntüler: - + Set Default Varsayılan olarak ayarla - + Settings Ayarlar - + Theme As in color theme Tema - + Language Dil - + Help Yardım - + Add Files... Dosya Ekle... - + Remove All Files Remove All Files - + Remove File Remove File - + Help Translate Panel Cleaner Help Translate Panel Cleaner - + Report an Issue... Report an Issue... - + Simulate Exception Simulate Exception - + Add File Browser Extension... Add File Browser Extension... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New Yeni - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Sil - + Import... İçe aktar... - + Online Documentation Çevrimiçi Dokümantasyon - + About Panel Cleaner Panel Temizleyici Hakkında - + View License Lisansı Görüntüle - + Save as... Farklı kaydet... - + System Sistem - + Dark Karanlık - + Light Açık - + Add Folder... Klasör Ekle... - + Delete Machine Learning Models Makine Öğrenimi Modellerini Sil - + Download Machine Learning Models Makine Öğrenimi Modellerini İndir - + Donate Bağış Yap @@ -2139,19 +2144,19 @@ Check the online documentation for how to add support for more languages.OCR modeli: - + Already downloaded Already downloaded - + Text Detector model (CUDA) Text Detector model (CUDA) - + Text Detector model (CPU) Text Detector model (CPU) @@ -2170,7 +2175,7 @@ Check the online documentation for how to add support for more languages.OCR download errors - + ETA estimated time of completion ETA @@ -2748,7 +2753,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2829,73 +2834,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input The original image with scaling applied (if needed). - + Input Step name in the image details view Input - + Input Output name in the image details view Input - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask The rough mask generated by the AI. - + Text Detection Step name in the image details view Text Detection - + AI Mask Output name in the image details view AI Mask - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes The unfiltered box data generated by the AI. - + Raw Boxes Output name in the image details view Raw Boxes - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes The outlines of the text boxes the AI found. - + Preprocessor Step name in the image details view Preprocessor - + Initial Boxes Output name in the image details view Initial Boxes - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2903,181 +2908,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. - + Final Boxes Output name in the image details view Final Boxes - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask The mask of the merged boxes. - + Masker Step name in the image details view Masker - + Box Mask Output name in the image details view Box Mask - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask The rough text detection mask with everything outside the box mask cut out. - + Cut Mask Output name in the image details view Cut Mask - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers The different steps of growth around the cut mask displayed in different colors. - + Mask Layers Output name in the image details view Mask Layers - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). - + Fitment Quality Output name in the image details view Fitment Quality - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay The input image with the final mask overlaid in color. - + Mask Overlay Output name in the image details view Mask Overlay - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask The collection of masks for each bubble that fit best. - + Final Mask Output name in the image details view Final Mask - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text The text layer isolated from the input image. - + Isolated Text Output name in the image details view Isolated Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output The input image with the final mask applied. - + Masked Output Output name in the image details view Masked Output - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask The masks that required denoising, to be overlaid on the final mask when exporting. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Denoise Mask - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output The input image with the denoised mask applied. - + Denoised Output Output name in the image details view Denoised Output - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask The inpainted sections around the text that was poorly cleaned, if at all. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Inpainted Mask - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output The input image: cleaned, denoised (if enabled), and inpainted. - + Inpainted Output Output name in the image details view Inpainted Output - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3086,1141 +3091,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title General - + About this profile: Profile comment in General About this profile: - + Notes Profile entry in General Notes - + Preferred file type to save the cleaned image as. Profile comment in General Preferred file type to save the cleaned image as. - + Preferred File Type Profile entry in General Preferred File Type - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Preferred file type to save the mask as. Only image formats that allow for transparency are supported. - + Preferred Mask File Type Profile entry in General Preferred Mask File Type - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. - + Input Height Lower Target Profile entry in General Input Height Lower Target - + Input Height Upper Target Profile entry in General Input Height Upper Target - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Text Detector - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> - + Model Path Profile entry in Text Detector Model Path - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! - + Concurrent Models Profile entry in Text Detector Concurrent Models - + Preprocessor Profile section title Preprocessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Minimum size of any box to keep it. - + Box Min Size Profile entry in Preprocessor Box Min Size - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. - + Suspicious Box Min Size Profile entry in Preprocessor Suspicious Box Min Size - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. - + Box Overlap Threshold Profile entry in Preprocessor Box Overlap Threshold - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. - + OCR Enabled Profile entry in Preprocessor OCR Enabled - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. - + OCR Use Tesseract Profile entry in Preprocessor OCR Use Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. - + OCR Engine Profile entry in Preprocessor OCR Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. - + Reading Order Profile entry in Preprocessor Reading Order - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. - + OCR Max Size Profile entry in Preprocessor OCR Max Size - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR Blacklist Pattern - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor OCR Strict Language - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. - + Box Padding Initial Profile entry in Preprocessor Box Padding Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. - + Box Right Padding Initial Profile entry in Preprocessor Box Right Padding Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Box Right Padding Extended - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. - + Box Reference Padding Profile entry in Preprocessor Box Reference Padding - + Masker Profile section title Masker - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. - + Mask Growth Step Pixels Profile entry in Masker Mask Growth Step Pixels - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. - + Mask Growth Steps Profile entry in Masker Mask Growth Steps - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. - + Min Mask Thickness Profile entry in Masker Min Mask Thickness - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Off White Max Threshold - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. - + Mask Improvement Threshold Profile entry in Masker Mask Improvement Threshold - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. - + Mask Selection Fast Profile entry in Masker Mask Selection Fast - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting Enabled - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. - + Min Inpainting Radius Profile entry in Inpainter Min Inpainting Radius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation - + Inpainting Radius Multiplier Profile entry in Inpainter Inpainting Radius Multiplier - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. - + Max Inpainting Radius Profile entry in Inpainter Max Inpainting Radius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. - + Inpainting Isolation Radius Profile entry in Inpainter Inpainting Isolation Radius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. - + Mask Max Standard Deviation Profile entry in Masker Mask Max Standard Deviation - + Color to use for the debug mask. Profile comment in Masker Color to use for the debug mask. - + Debug Mask Color Profile entry in Masker Debug Mask Color - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. - + Denoising Enabled Profile entry in Denoiser Denoising Enabled - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. - + Noise Min Standard Deviation Profile entry in Denoiser Noise Min Standard Deviation - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser The thickness of an outline to denoise around a mask. - + Noise Outline Size Profile entry in Denoiser Noise Outline Size - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. - + Noise Fade Radius Profile entry in Denoiser Noise Fade Radius - + Set to True to support colored images. Profile comment in Denoiser Set to True to support colored images. - + Colored Images Profile entry in Denoiser Colored Images - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the image. Higher values mean more denoising. - + Filter Strength Profile entry in Denoiser Filter Strength - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. - + Color Filter Strength Profile entry in Denoiser Color Filter Strength - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Size in pixels of the template patch that is used to compute weights. Should be odd. - + Template Window Size Profile entry in Denoiser Template Window Size - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. - + Search Window Size Profile entry in Denoiser Search Window Size @@ -4247,7 +4252,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4664,14 +4669,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From 58be958b5f1a0ec72c977bea3c05604f65fe9e9b Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:22 +0100 Subject: [PATCH 15/18] New translations panelcleaner_source.ts (Ukrainian) --- translations/PanelCleaner_uk_UA.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_uk_UA.ts b/translations/PanelCleaner_uk_UA.ts index 679c9a4..8a2d700 100644 --- a/translations/PanelCleaner_uk_UA.ts +++ b/translations/PanelCleaner_uk_UA.ts @@ -4,7 +4,7 @@ - + default reserved profile name default @@ -20,54 +20,54 @@ Profile {profile_name} created. - + Profile name cannot be empty. Profile name cannot be empty. - + Profile name already in use. Profile name already in use. - + Profile name is reserved. Profile name is reserved. - - - - + + + + Unknown Unknown - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Grayscale - + 1-bit Color mode 1-bit - + Palette Color mode Palette @@ -78,45 +78,45 @@ Error downloading file from url: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Error downloading file from url: {url} Failed to save the file to {save_path} - + Error downloading file from url: {url} The file content is different from expected. Error downloading file from url: {url} The file content is different from expected. + - - - + + + - - - + + N/A N/A - + OCR Analytics OCR Analytics - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) @@ -155,66 +155,66 @@ Removed bubbles: - + Small boxes Small boxes - + Removed boxes Removed boxes - - + + Failed Failed - + Perfect Perfect - + Total Total - + Mask Fitment Analytics Mask Fitment Analytics - + Total Total - + Total boxes Total boxes - + Masks succeeded Masks succeeded - + Masks failed Masks failed - + Perfect masks Perfect masks - + Average border deviation Average border deviation @@ -233,7 +233,7 @@ Mask usage by thickness (in pixels): Mask ({thickness}px) - + Box mask Box mask @@ -254,27 +254,27 @@ Pages with failures / total: - + Denoising Analytics Denoising Analytics - + Total masks Total masks - + Masks denoised Masks denoised - + Minimum deviation to denoise Minimum deviation to denoise - + Maximum allowed deviation Maximum allowed deviation @@ -293,27 +293,27 @@ Pages with failures / total: Denoised - + Inpainting Analytics Inpainting Analytics - + Inpainting performed Inpainting performed - + Average thickness Average thickness - + Minimum thickness Minimum thickness - + Maximum thickness Maximum thickness @@ -372,7 +372,7 @@ Pages with failures / total: The file format was not recognized. - + The CSV file must start with a header row, followed by data rows. The CSV file must start with a header row, followed by data rows. @@ -402,7 +402,7 @@ Pages with failures / total: Multiple images match the file path: {path} - + Error, no version found. Error, no version found. @@ -412,7 +412,7 @@ Pages with failures / total: CUDA Error - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pages with failures / total: Inpainting model not found. - + filename filename - + startx startx - + starty starty - + endx endx - + endy endy - + text text - + Saved detected text to {output_file} Saved detected text to {output_file} - + Failed to write detected text to {output_file} Failed to write detected text to {output_file} - + Save Failed Save Failed - + Failed to write detected text to file. Failed to write detected text to file. @@ -546,7 +546,7 @@ Pages with failures / total: - + Page layered export @@ -635,7 +635,7 @@ An AI-powered tool to clean manga panels. ColorButton - + Select Color Select Color @@ -681,7 +681,7 @@ An AI-powered tool to clean manga panels. Unsupported system - + This system is not supported. This system is not supported. @@ -696,12 +696,12 @@ An AI-powered tool to clean manga panels. Uninstall for Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. The extension will be installed at {path}. This doesn't require root permissions. - + The extension file will be deleted. This doesn't require root permissions. The extension file will be deleted. This doesn't require root permissions. @@ -716,17 +716,17 @@ An AI-powered tool to clean manga panels. Uninstall for Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. The extension will be installed in the Windows registry.This requires admin permissions. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. The extension will be uninstalled from the Windows registry.This requires admin permissions. - + Dolphin extension installed successfully. Dolphin extension installed successfully. @@ -736,7 +736,7 @@ An AI-powered tool to clean manga panels. Installation successful - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Dolphin extension installed successfully. @@ -816,12 +816,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) Original size in pixels (width × height) - + Processing size in pixels (width × height), scale factor Processing size in pixels (width × height), scale factor @@ -847,7 +847,7 @@ By default, this uses the built-in profile. However, you can change this default Unsupported TIFF files - + The following 5-channel TIFF files are not supported: The following 5-channel TIFF files are not supported: @@ -874,27 +874,27 @@ By default, this uses the built-in profile. However, you can change this default split - + Number of boxes removed by the OCR model / total boxes Number of boxes removed by the OCR model / total boxes - + Number of boxes that failed to generate a mask / total boxes Number of boxes that failed to generate a mask / total boxes - + Number of boxes that were perfectly masked / total boxes Number of boxes that were perfectly masked / total boxes - + Number of boxes that were denoised / total boxes Number of boxes that were denoised / total boxes - + Number of boxes that were inpainted Number of boxes that were inpainted @@ -1042,14 +1042,14 @@ By default, this uses the built-in profile. However, you can change this default Impossible Request - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting is disabled in the current profile, this output can't be generated. @@ -1071,7 +1071,7 @@ Please enable inpainting in the profile settings and try again. Export failed - + Failed to export image: Failed to export image: @@ -1114,7 +1114,7 @@ Please enable inpainting in the profile settings and try again. result - + results results @@ -1129,17 +1129,17 @@ Please enable inpainting in the profile settings and try again. new image - + new images new images - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. @@ -1238,12 +1238,12 @@ Please enable inpainting in the profile settings and try again. Note: Name "{name}" was hidden - + Corrupted log session Corrupted log session - + Current session Current session @@ -1258,7 +1258,7 @@ Please enable inpainting in the profile settings and try again. Error - + Errors Errors @@ -1268,7 +1268,7 @@ Please enable inpainting in the profile settings and try again. Critical - + Criticals Criticals @@ -1284,49 +1284,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button Open Config - + Open Log Statusbar button Open Log - + CUDA Enabled CUDA Enabled - + Select Output Directory Select Output Directory - + Aborting... Aborting... - + Loading OCR model... Loading OCR model... - + Loaded OCR model. Loaded OCR model. - + Error Error - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1335,7 +1335,7 @@ Continue anyway? Continue anyway? - + Delete Models Delete Models @@ -1345,23 +1345,23 @@ Continue anyway? Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. - + Encountered error: Encountered error: - + MPS Enabled MPS Enabled - - + + Inpainting Model Update Inpainting Model Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1375,148 +1375,148 @@ Switch to the new model? Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. - + Layout Reset Layout Reset - + The window layout has been reset. The default layout will be used next time. The window layout has been reset. The default layout will be used next time. - + Multiple Instances Multiple Instances - + Found process ID in lock file: Found process ID in lock file: - + Failed to delete models. Failed to delete models. - + Delete Failed Delete Failed - + Models Deleted Models Deleted - + The models were deleted. Would you like to download them again? The models were deleted. Would you like to download them again? - + Not Supported Not Supported - + Unsupported system or package format Unsupported system or package format - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Restart Required - + The language has been changed. Please restart the application for the changes to take effect. The language has been changed. Please restart the application for the changes to take effect. - - + + System Language System Language - + Import Profile Import Profile - + Profile Files (*.conf) Profile Files (*.conf) - + Profile Imported Profile Imported - + Import Error Import Error - + Failed to Delete Failed to Delete - + The default profile cannot be deleted. The default profile cannot be deleted. - + Delete Profile Delete Profile - + Are you sure you want to delete the profile {profile_name}? Are you sure you want to delete the profile {profile_name}? - + Delete Error Delete Error - + Failed to delete the profile. Failed to delete the profile. @@ -1528,217 +1528,217 @@ Switching profiles will discard changes to the current profile. Switching profiles will discard changes to the current profile. - + Unsaved changes Unsaved changes - - + + Load Error Load Error - - + + Failed to load profile: Failed to load profile: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Profile Created - + Create Error Create Error - + Failed to save profile to {profile_path} Failed to save profile to {profile_path} - - + + Save Error Save Error - + Failed to save profile. Failed to save profile. - + Profile saved to {profile_path} Profile saved to {profile_path} - + Failed to save config. Failed to save config. - + Failed to save the new profile to the configuration file. Failed to save the new profile to the configuration file. - - + + No Files No Files - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. - + No Outputs No Outputs - + No outputs were requested. Please select at least one output before cleaning. No outputs were requested. Please select at least one output before cleaning. - - + + cleaned cleaned - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists File Exists - + The file '{output_path}' already exists. Overwrite? The file '{output_path}' already exists. Overwrite? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Processing Finished - + Finished processing all files. Finished processing all files. - + Processing Aborted Processing Aborted - + Processing aborted. Processing aborted. - + Processing Error Processing Error - + Encountered an error while processing files. Encountered an error while processing files. @@ -1749,7 +1749,7 @@ Check the online documentation for how to add support for more languages. - + Profile Profile @@ -1764,360 +1764,365 @@ Check the online documentation for how to add support for more languages.Apply - - + + Save Save - + Reset All Reset All - + Drag and Drop Images or Folders Here Drag and Drop Images or Folders Here - + Images Images - - + + File File - + Size Size - + Processing Size The size that the picture is shrunk down to for processing. Processing Size - + File Size File Size - + Color Mode Like RGB, grayscale etc. Color Mode - + Analytics Can also call this statistics. Analytics - + Process Process - + Cleaning Cleaning - + OCR OCR - + Edit Existing Output Edit Existing Output - + Abort Abort - + Start Start - + Output Output - + Cleaned Image Cleaned Image - + Mask Mask - + Isolated Text Isolated Text - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Save Output - + Output Directory: Output Directory: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. You can use a relative path to create a subfolder at the image's original location, or use an absolute path. - + Plain Text Plain Text - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Output File: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Generating: - + Current Step: Current Step: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Images Processed: - + Set Default Set Default - + Settings Settings - + Theme As in color theme Theme - + Language Language - + Help Help - + Add Files... Add Files... - + Remove All Files Remove All Files - + Remove File Remove File - + Help Translate Panel Cleaner Help Translate Panel Cleaner - + Report an Issue... Report an Issue... - + Simulate Exception Simulate Exception - + Add File Browser Extension... Add File Browser Extension... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New New - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Delete - + Import... Import... - + Online Documentation Online Documentation - + About Panel Cleaner About Panel Cleaner - + View License View License - + Save as... Save as... - + System System - + Dark Dark - + Light Light - + Add Folder... Add Folder... - + Delete Machine Learning Models Delete Machine Learning Models - + Download Machine Learning Models Download Machine Learning Models - + Donate Donate @@ -2149,19 +2154,19 @@ Check the online documentation for how to add support for more languages.OCR model: - + Already downloaded Already downloaded - + Text Detector model (CUDA) Text Detector model (CUDA) - + Text Detector model (CPU) Text Detector model (CPU) @@ -2180,7 +2185,7 @@ Check the online documentation for how to add support for more languages.OCR download errors - + ETA estimated time of completion ETA @@ -2758,7 +2763,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2839,73 +2844,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input The original image with scaling applied (if needed). - + Input Step name in the image details view Input - + Input Output name in the image details view Input - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask The rough mask generated by the AI. - + Text Detection Step name in the image details view Text Detection - + AI Mask Output name in the image details view AI Mask - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes The unfiltered box data generated by the AI. - + Raw Boxes Output name in the image details view Raw Boxes - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes The outlines of the text boxes the AI found. - + Preprocessor Step name in the image details view Preprocessor - + Initial Boxes Output name in the image details view Initial Boxes - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2913,181 +2918,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. - + Final Boxes Output name in the image details view Final Boxes - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask The mask of the merged boxes. - + Masker Step name in the image details view Masker - + Box Mask Output name in the image details view Box Mask - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask The rough text detection mask with everything outside the box mask cut out. - + Cut Mask Output name in the image details view Cut Mask - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers The different steps of growth around the cut mask displayed in different colors. - + Mask Layers Output name in the image details view Mask Layers - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). - + Fitment Quality Output name in the image details view Fitment Quality - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay The input image with the final mask overlaid in color. - + Mask Overlay Output name in the image details view Mask Overlay - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask The collection of masks for each bubble that fit best. - + Final Mask Output name in the image details view Final Mask - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text The text layer isolated from the input image. - + Isolated Text Output name in the image details view Isolated Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output The input image with the final mask applied. - + Masked Output Output name in the image details view Masked Output - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask The masks that required denoising, to be overlaid on the final mask when exporting. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Denoise Mask - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output The input image with the denoised mask applied. - + Denoised Output Output name in the image details view Denoised Output - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask The inpainted sections around the text that was poorly cleaned, if at all. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Inpainted Mask - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output The input image: cleaned, denoised (if enabled), and inpainted. - + Inpainted Output Output name in the image details view Inpainted Output - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3096,1141 +3101,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title General - + About this profile: Profile comment in General About this profile: - + Notes Profile entry in General Notes - + Preferred file type to save the cleaned image as. Profile comment in General Preferred file type to save the cleaned image as. - + Preferred File Type Profile entry in General Preferred File Type - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Preferred file type to save the mask as. Only image formats that allow for transparency are supported. - + Preferred Mask File Type Profile entry in General Preferred Mask File Type - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. - + Input Height Lower Target Profile entry in General Input Height Lower Target - + Input Height Upper Target Profile entry in General Input Height Upper Target - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Text Detector - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> - + Model Path Profile entry in Text Detector Model Path - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! - + Concurrent Models Profile entry in Text Detector Concurrent Models - + Preprocessor Profile section title Preprocessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Minimum size of any box to keep it. - + Box Min Size Profile entry in Preprocessor Box Min Size - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. - + Suspicious Box Min Size Profile entry in Preprocessor Suspicious Box Min Size - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. - + Box Overlap Threshold Profile entry in Preprocessor Box Overlap Threshold - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. - + OCR Enabled Profile entry in Preprocessor OCR Enabled - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. - + OCR Use Tesseract Profile entry in Preprocessor OCR Use Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. - + OCR Engine Profile entry in Preprocessor OCR Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. - + Reading Order Profile entry in Preprocessor Reading Order - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. - + OCR Max Size Profile entry in Preprocessor OCR Max Size - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR Blacklist Pattern - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor OCR Strict Language - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. - + Box Padding Initial Profile entry in Preprocessor Box Padding Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. - + Box Right Padding Initial Profile entry in Preprocessor Box Right Padding Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Box Right Padding Extended - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. - + Box Reference Padding Profile entry in Preprocessor Box Reference Padding - + Masker Profile section title Masker - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. - + Mask Growth Step Pixels Profile entry in Masker Mask Growth Step Pixels - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. - + Mask Growth Steps Profile entry in Masker Mask Growth Steps - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. - + Min Mask Thickness Profile entry in Masker Min Mask Thickness - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Off White Max Threshold - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. - + Mask Improvement Threshold Profile entry in Masker Mask Improvement Threshold - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. - + Mask Selection Fast Profile entry in Masker Mask Selection Fast - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting Enabled - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. - + Min Inpainting Radius Profile entry in Inpainter Min Inpainting Radius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation - + Inpainting Radius Multiplier Profile entry in Inpainter Inpainting Radius Multiplier - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. - + Max Inpainting Radius Profile entry in Inpainter Max Inpainting Radius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. - + Inpainting Isolation Radius Profile entry in Inpainter Inpainting Isolation Radius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. - + Mask Max Standard Deviation Profile entry in Masker Mask Max Standard Deviation - + Color to use for the debug mask. Profile comment in Masker Color to use for the debug mask. - + Debug Mask Color Profile entry in Masker Debug Mask Color - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. - + Denoising Enabled Profile entry in Denoiser Denoising Enabled - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. - + Noise Min Standard Deviation Profile entry in Denoiser Noise Min Standard Deviation - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser The thickness of an outline to denoise around a mask. - + Noise Outline Size Profile entry in Denoiser Noise Outline Size - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. - + Noise Fade Radius Profile entry in Denoiser Noise Fade Radius - + Set to True to support colored images. Profile comment in Denoiser Set to True to support colored images. - + Colored Images Profile entry in Denoiser Colored Images - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the image. Higher values mean more denoising. - + Filter Strength Profile entry in Denoiser Filter Strength - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. - + Color Filter Strength Profile entry in Denoiser Color Filter Strength - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Size in pixels of the template patch that is used to compute weights. Should be odd. - + Template Window Size Profile entry in Denoiser Template Window Size - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. - + Search Window Size Profile entry in Denoiser Search Window Size @@ -4257,7 +4262,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4674,14 +4679,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From 9358163bba6146062ee6b2c00f5dbe595d6b5a3f Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:23 +0100 Subject: [PATCH 16/18] New translations panelcleaner_source.ts (Chinese Simplified) --- translations/PanelCleaner_zh_CN.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_zh_CN.ts b/translations/PanelCleaner_zh_CN.ts index a7aa1ba..1e9e3f6 100644 --- a/translations/PanelCleaner_zh_CN.ts +++ b/translations/PanelCleaner_zh_CN.ts @@ -4,7 +4,7 @@ - + default reserved profile name default @@ -20,54 +20,54 @@ Profile {profile_name} created. - + Profile name cannot be empty. Profile name cannot be empty. - + Profile name already in use. Profile name already in use. - + Profile name is reserved. Profile name is reserved. - - - - + + + + Unknown Unknown - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Grayscale - + 1-bit Color mode 1-bit - + Palette Color mode Palette @@ -78,45 +78,45 @@ Error downloading file from url: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Error downloading file from url: {url} Failed to save the file to {save_path} - + Error downloading file from url: {url} The file content is different from expected. Error downloading file from url: {url} The file content is different from expected. + - - - + + + - - - + + N/A N/A - + OCR Analytics OCR Analytics - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) @@ -155,66 +155,66 @@ Removed bubbles: - + Small boxes Small boxes - + Removed boxes Removed boxes - - + + Failed Failed - + Perfect Perfect - + Total Total - + Mask Fitment Analytics Mask Fitment Analytics - + Total Total - + Total boxes Total boxes - + Masks succeeded Masks succeeded - + Masks failed Masks failed - + Perfect masks Perfect masks - + Average border deviation Average border deviation @@ -233,7 +233,7 @@ Mask usage by thickness (in pixels): Mask ({thickness}px) - + Box mask Box mask @@ -254,27 +254,27 @@ Pages with failures / total: - + Denoising Analytics Denoising Analytics - + Total masks Total masks - + Masks denoised Masks denoised - + Minimum deviation to denoise Minimum deviation to denoise - + Maximum allowed deviation Maximum allowed deviation @@ -293,27 +293,27 @@ Pages with failures / total: Denoised - + Inpainting Analytics Inpainting Analytics - + Inpainting performed Inpainting performed - + Average thickness Average thickness - + Minimum thickness Minimum thickness - + Maximum thickness Maximum thickness @@ -372,7 +372,7 @@ Pages with failures / total: The file format was not recognized. - + The CSV file must start with a header row, followed by data rows. The CSV file must start with a header row, followed by data rows. @@ -402,7 +402,7 @@ Pages with failures / total: Multiple images match the file path: {path} - + Error, no version found. Error, no version found. @@ -412,7 +412,7 @@ Pages with failures / total: CUDA Error - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> @@ -422,61 +422,61 @@ Pages with failures / total: Inpainting model not found. - + filename filename - + startx startx - + starty starty - + endx endx - + endy endy - + text text - + Saved detected text to {output_file} Saved detected text to {output_file} - + Failed to write detected text to {output_file} Failed to write detected text to {output_file} - + Save Failed Save Failed - + Failed to write detected text to file. Failed to write detected text to file. @@ -534,7 +534,7 @@ Pages with failures / total: - + Page layered export @@ -620,7 +620,7 @@ An AI-powered tool to clean manga panels. ColorButton - + Select Color Select Color @@ -666,7 +666,7 @@ An AI-powered tool to clean manga panels. Unsupported system - + This system is not supported. This system is not supported. @@ -681,12 +681,12 @@ An AI-powered tool to clean manga panels. Uninstall for Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. The extension will be installed at {path}. This doesn't require root permissions. - + The extension file will be deleted. This doesn't require root permissions. The extension file will be deleted. This doesn't require root permissions. @@ -701,17 +701,17 @@ An AI-powered tool to clean manga panels. Uninstall for Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. The extension will be installed in the Windows registry.This requires admin permissions. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. The extension will be uninstalled from the Windows registry.This requires admin permissions. - + Dolphin extension installed successfully. Dolphin extension installed successfully. @@ -721,7 +721,7 @@ An AI-powered tool to clean manga panels. Installation successful - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. Dolphin extension installed successfully. @@ -801,12 +801,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) Original size in pixels (width × height) - + Processing size in pixels (width × height), scale factor Processing size in pixels (width × height), scale factor @@ -832,7 +832,7 @@ By default, this uses the built-in profile. However, you can change this default Unsupported TIFF files - + The following 5-channel TIFF files are not supported: The following 5-channel TIFF files are not supported: @@ -859,27 +859,27 @@ By default, this uses the built-in profile. However, you can change this default split - + Number of boxes removed by the OCR model / total boxes Number of boxes removed by the OCR model / total boxes - + Number of boxes that failed to generate a mask / total boxes Number of boxes that failed to generate a mask / total boxes - + Number of boxes that were perfectly masked / total boxes Number of boxes that were perfectly masked / total boxes - + Number of boxes that were denoised / total boxes Number of boxes that were denoised / total boxes - + Number of boxes that were inpainted Number of boxes that were inpainted @@ -1027,14 +1027,14 @@ By default, this uses the built-in profile. However, you can change this default Impossible Request - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting is disabled in the current profile, this output can't be generated. @@ -1056,7 +1056,7 @@ Please enable inpainting in the profile settings and try again. Export failed - + Failed to export image: Failed to export image: @@ -1099,7 +1099,7 @@ Please enable inpainting in the profile settings and try again. result - + results results @@ -1114,17 +1114,17 @@ Please enable inpainting in the profile settings and try again. new image - + new images new images - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. @@ -1223,12 +1223,12 @@ Please enable inpainting in the profile settings and try again. Note: Name "{name}" was hidden - + Corrupted log session Corrupted log session - + Current session Current session @@ -1243,7 +1243,7 @@ Please enable inpainting in the profile settings and try again. Error - + Errors Errors @@ -1253,7 +1253,7 @@ Please enable inpainting in the profile settings and try again. Critical - + Criticals Criticals @@ -1269,49 +1269,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button Open Config - + Open Log Statusbar button Open Log - + CUDA Enabled CUDA Enabled - + Select Output Directory Select Output Directory - + Aborting... Aborting... - + Loading OCR model... Loading OCR model... - + Loaded OCR model. Loaded OCR model. - + Error Error - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1320,7 +1320,7 @@ Continue anyway? Continue anyway? - + Delete Models Delete Models @@ -1330,23 +1330,23 @@ Continue anyway? Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. - + Encountered error: Encountered error: - + MPS Enabled MPS Enabled - - + + Inpainting Model Update Inpainting Model Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1360,148 +1360,148 @@ Switch to the new model? Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. - + Layout Reset Layout Reset - + The window layout has been reset. The default layout will be used next time. The window layout has been reset. The default layout will be used next time. - + Multiple Instances Multiple Instances - + Found process ID in lock file: Found process ID in lock file: - + Failed to delete models. Failed to delete models. - + Delete Failed Delete Failed - + Models Deleted Models Deleted - + The models were deleted. Would you like to download them again? The models were deleted. Would you like to download them again? - + Not Supported Not Supported - + Unsupported system or package format Unsupported system or package format - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Restart Required - + The language has been changed. Please restart the application for the changes to take effect. The language has been changed. Please restart the application for the changes to take effect. - - + + System Language System Language - + Import Profile Import Profile - + Profile Files (*.conf) Profile Files (*.conf) - + Profile Imported Profile Imported - + Import Error Import Error - + Failed to Delete Failed to Delete - + The default profile cannot be deleted. The default profile cannot be deleted. - + Delete Profile Delete Profile - + Are you sure you want to delete the profile {profile_name}? Are you sure you want to delete the profile {profile_name}? - + Delete Error Delete Error - + Failed to delete the profile. Failed to delete the profile. @@ -1513,217 +1513,217 @@ Switching profiles will discard changes to the current profile. Switching profiles will discard changes to the current profile. - + Unsaved changes Unsaved changes - - + + Load Error Load Error - - + + Failed to load profile: Failed to load profile: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Profile Created - + Create Error Create Error - + Failed to save profile to {profile_path} Failed to save profile to {profile_path} - - + + Save Error Save Error - + Failed to save profile. Failed to save profile. - + Profile saved to {profile_path} Profile saved to {profile_path} - + Failed to save config. Failed to save config. - + Failed to save the new profile to the configuration file. Failed to save the new profile to the configuration file. - - + + No Files No Files - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. - + No Outputs No Outputs - + No outputs were requested. Please select at least one output before cleaning. No outputs were requested. Please select at least one output before cleaning. - - + + cleaned cleaned - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists File Exists - + The file '{output_path}' already exists. Overwrite? The file '{output_path}' already exists. Overwrite? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Processing Finished - + Finished processing all files. Finished processing all files. - + Processing Aborted Processing Aborted - + Processing aborted. Processing aborted. - + Processing Error Processing Error - + Encountered an error while processing files. Encountered an error while processing files. @@ -1734,7 +1734,7 @@ Check the online documentation for how to add support for more languages. - + Profile Profile @@ -1749,360 +1749,365 @@ Check the online documentation for how to add support for more languages.Apply - - + + Save Save - + Reset All Reset All - + Drag and Drop Images or Folders Here Drag and Drop Images or Folders Here - + Images Images - - + + File File - + Size Size - + Processing Size The size that the picture is shrunk down to for processing. Processing Size - + File Size File Size - + Color Mode Like RGB, grayscale etc. Color Mode - + Analytics Can also call this statistics. Analytics - + Process Process - + Cleaning Cleaning - + OCR OCR - + Edit Existing Output Edit Existing Output - + Abort Abort - + Start Start - + Output Output - + Cleaned Image Cleaned Image - + Mask Mask - + Isolated Text Isolated Text - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Save Output - + Output Directory: Output Directory: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. You can use a relative path to create a subfolder at the image's original location, or use an absolute path. - + Plain Text Plain Text - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Output File: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Generating: - + Current Step: Current Step: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Images Processed: - + Set Default Set Default - + Settings Settings - + Theme As in color theme Theme - + Language Language - + Help Help - + Add Files... Add Files... - + Remove All Files Remove All Files - + Remove File Remove File - + Help Translate Panel Cleaner Help Translate Panel Cleaner - + Report an Issue... Report an Issue... - + Simulate Exception Simulate Exception - + Add File Browser Extension... Add File Browser Extension... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New New - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Delete - + Import... Import... - + Online Documentation Online Documentation - + About Panel Cleaner About Panel Cleaner - + View License View License - + Save as... Save as... - + System System - + Dark Dark - + Light Light - + Add Folder... Add Folder... - + Delete Machine Learning Models Delete Machine Learning Models - + Download Machine Learning Models Download Machine Learning Models - + Donate Donate @@ -2134,19 +2139,19 @@ Check the online documentation for how to add support for more languages.OCR model: - + Already downloaded Already downloaded - + Text Detector model (CUDA) Text Detector model (CUDA) - + Text Detector model (CPU) Text Detector model (CPU) @@ -2165,7 +2170,7 @@ Check the online documentation for how to add support for more languages.OCR download errors - + ETA estimated time of completion ETA @@ -2743,7 +2748,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2824,73 +2829,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input The original image with scaling applied (if needed). - + Input Step name in the image details view Input - + Input Output name in the image details view Input - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask The rough mask generated by the AI. - + Text Detection Step name in the image details view Text Detection - + AI Mask Output name in the image details view AI Mask - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes The unfiltered box data generated by the AI. - + Raw Boxes Output name in the image details view Raw Boxes - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes The outlines of the text boxes the AI found. - + Preprocessor Step name in the image details view Preprocessor - + Initial Boxes Output name in the image details view Initial Boxes - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2898,181 +2903,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. - + Final Boxes Output name in the image details view Final Boxes - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask The mask of the merged boxes. - + Masker Step name in the image details view Masker - + Box Mask Output name in the image details view Box Mask - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask The rough text detection mask with everything outside the box mask cut out. - + Cut Mask Output name in the image details view Cut Mask - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers The different steps of growth around the cut mask displayed in different colors. - + Mask Layers Output name in the image details view Mask Layers - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). - + Fitment Quality Output name in the image details view Fitment Quality - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay The input image with the final mask overlaid in color. - + Mask Overlay Output name in the image details view Mask Overlay - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask The collection of masks for each bubble that fit best. - + Final Mask Output name in the image details view Final Mask - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text The text layer isolated from the input image. - + Isolated Text Output name in the image details view Isolated Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output The input image with the final mask applied. - + Masked Output Output name in the image details view Masked Output - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask The masks that required denoising, to be overlaid on the final mask when exporting. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Denoise Mask - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output The input image with the denoised mask applied. - + Denoised Output Output name in the image details view Denoised Output - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask The inpainted sections around the text that was poorly cleaned, if at all. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Inpainted Mask - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output The input image: cleaned, denoised (if enabled), and inpainted. - + Inpainted Output Output name in the image details view Inpainted Output - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3081,1141 +3086,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title General - + About this profile: Profile comment in General About this profile: - + Notes Profile entry in General Notes - + Preferred file type to save the cleaned image as. Profile comment in General Preferred file type to save the cleaned image as. - + Preferred File Type Profile entry in General Preferred File Type - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Preferred file type to save the mask as. Only image formats that allow for transparency are supported. - + Preferred Mask File Type Profile entry in General Preferred Mask File Type - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. - + Input Height Lower Target Profile entry in General Input Height Lower Target - + Input Height Upper Target Profile entry in General Input Height Upper Target - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Text Detector - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> - + Model Path Profile entry in Text Detector Model Path - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! - + Concurrent Models Profile entry in Text Detector Concurrent Models - + Preprocessor Profile section title Preprocessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Minimum size of any box to keep it. - + Box Min Size Profile entry in Preprocessor Box Min Size - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. - + Suspicious Box Min Size Profile entry in Preprocessor Suspicious Box Min Size - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. - + Box Overlap Threshold Profile entry in Preprocessor Box Overlap Threshold - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. - + OCR Enabled Profile entry in Preprocessor OCR Enabled - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. - + OCR Use Tesseract Profile entry in Preprocessor OCR Use Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. - + OCR Engine Profile entry in Preprocessor OCR Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. - + Reading Order Profile entry in Preprocessor Reading Order - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. - + OCR Max Size Profile entry in Preprocessor OCR Max Size - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR Blacklist Pattern - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor OCR Strict Language - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. - + Box Padding Initial Profile entry in Preprocessor Box Padding Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. - + Box Right Padding Initial Profile entry in Preprocessor Box Right Padding Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Box Right Padding Extended - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. - + Box Reference Padding Profile entry in Preprocessor Box Reference Padding - + Masker Profile section title Masker - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. - + Mask Growth Step Pixels Profile entry in Masker Mask Growth Step Pixels - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. - + Mask Growth Steps Profile entry in Masker Mask Growth Steps - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. - + Min Mask Thickness Profile entry in Masker Min Mask Thickness - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Off White Max Threshold - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. - + Mask Improvement Threshold Profile entry in Masker Mask Improvement Threshold - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. - + Mask Selection Fast Profile entry in Masker Mask Selection Fast - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting Enabled - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. - + Min Inpainting Radius Profile entry in Inpainter Min Inpainting Radius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation - + Inpainting Radius Multiplier Profile entry in Inpainter Inpainting Radius Multiplier - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. - + Max Inpainting Radius Profile entry in Inpainter Max Inpainting Radius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. - + Inpainting Isolation Radius Profile entry in Inpainter Inpainting Isolation Radius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. - + Mask Max Standard Deviation Profile entry in Masker Mask Max Standard Deviation - + Color to use for the debug mask. Profile comment in Masker Color to use for the debug mask. - + Debug Mask Color Profile entry in Masker Debug Mask Color - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. - + Denoising Enabled Profile entry in Denoiser Denoising Enabled - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. - + Noise Min Standard Deviation Profile entry in Denoiser Noise Min Standard Deviation - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser The thickness of an outline to denoise around a mask. - + Noise Outline Size Profile entry in Denoiser Noise Outline Size - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. - + Noise Fade Radius Profile entry in Denoiser Noise Fade Radius - + Set to True to support colored images. Profile comment in Denoiser Set to True to support colored images. - + Colored Images Profile entry in Denoiser Colored Images - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the image. Higher values mean more denoising. - + Filter Strength Profile entry in Denoiser Filter Strength - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. - + Color Filter Strength Profile entry in Denoiser Color Filter Strength - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Size in pixels of the template patch that is used to compute weights. Should be odd. - + Template Window Size Profile entry in Denoiser Template Window Size - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. - + Search Window Size Profile entry in Denoiser Search Window Size @@ -4242,7 +4247,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4659,14 +4664,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default. From c8770aabbd86457ce8bcfd3070a02d53b3aacfb7 Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:24 +0100 Subject: [PATCH 17/18] New translations panelcleaner_source.ts (Portuguese, Brazilian) --- translations/PanelCleaner_pt_BR.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_pt_BR.ts b/translations/PanelCleaner_pt_BR.ts index c81b70b..0ee0ee7 100644 --- a/translations/PanelCleaner_pt_BR.ts +++ b/translations/PanelCleaner_pt_BR.ts @@ -4,7 +4,7 @@ - + default reserved profile name padrão @@ -20,54 +20,54 @@ Perfil {profile_name} criado. - + Profile name cannot be empty. O nome do perfil não pode estar branco. - + Profile name already in use. Nome do perfil já em uso. - + Profile name is reserved. O nome do perfil é reservado. - - - - + + + + Unknown Desconhecido - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode Tons de cinza - + 1-bit Color mode 1 bit - + Palette Color mode Paleta @@ -78,45 +78,45 @@ Erro ao baixar o arquivo de URL: {url} - + Error downloading file from url: {url} Failed to save the file to {save_path} Erro ao baixar o arquivo de (URL): {url} Falha ao salvar o arquivo em {save_path} - + Error downloading file from url: {url} The file content is different from expected. Erro ao baixar o arquivo da (URL): {url} O conteúdo do arquivo difere do esperado. + - - - + + + - - - + + N/A N/A - + OCR Analytics Análise de OCR - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) Número de quadrados: {num_boxes} | Número de balões pequenos: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) Número de quadrados removidos: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} de quadrados pequenos) @@ -152,65 +152,65 @@ Removed bubbles: - + Small boxes Quadrados pequenos - + Removed boxes Quadrados removidos - - + + Failed Falha - + Perfect Perfeito - + Total Total - + Mask Fitment Analytics Análise de ajuste da máscara - + Total Total - + Total boxes Total de quadrados - + Masks succeeded Máscaras bem-sucedidas - + Masks failed Máscaras fracassadas - + Perfect masks Máscaras perfeitas - + Average border deviation Desvio médio da borda @@ -229,7 +229,7 @@ Uso da máscara por espessura (em pixels): Máscara ({thickness}px) - + Box mask Máscara de quadrado @@ -250,27 +250,27 @@ Páginas com falhas / total: - + Denoising Analytics Análise de tratamento (Denoise) - + Total masks Total de máscaras - + Masks denoised Máscaras sem ruído - + Minimum deviation to denoise Desvio mínimo para tratar (Denoise) - + Maximum allowed deviation Desvio máximo permitido @@ -289,27 +289,27 @@ Páginas com falhas / total: Tratado (Denoised) - + Inpainting Analytics Análise de Inpainting - + Inpainting performed Inpaint feita - + Average thickness Espessura média - + Minimum thickness Espessura mínima - + Maximum thickness Espessura máxima @@ -368,7 +368,7 @@ Páginas com falhas / total: O formato do arquivo não foi reconhecido. - + The CSV file must start with a header row, followed by data rows. O arquivo CSV deve começar com uma linha de cabeçalho, seguida pelas linhas de dados. @@ -398,7 +398,7 @@ Páginas com falhas / total: Várias imagens correspondem ao caminho do arquivo: {path} - + Error, no version found. Erro, nenhuma versão encontrada. @@ -408,7 +408,7 @@ Páginas com falhas / total: Erro CUDA - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>Sua GPU não é compatível com as operações CUDA necessárias.<br><br>Tente desinstalar as versões atuais do torch e do torchvision e instalar a versão da CPU (ou uma versão CUDA diferente).<br>Você pode encontrar mais instruções aqui: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Verifique a seção "Compute Platform" para ver as versões disponíveis.<br><br>Sua versão atual do CUDA é: {cuda_version}<br></html> @@ -418,61 +418,61 @@ Páginas com falhas / total: Modelo de pintura não encontrado. - + filename nome do arquivo - + startx startx - + starty starty - + endx endx - + endy endy - + text texto - + Saved detected text to {output_file} Texto detectado salvo em {output_file} - + Failed to write detected text to {output_file} Falha ao gravar o texto detectado em {output_file} - + Save Failed Falha ao salvar - + Failed to write detected text to file. Falha ao gravar o texto detectado no arquivo. @@ -534,7 +534,7 @@ Páginas com falhas / total: - + Page layered export @@ -621,7 +621,7 @@ Uma ferramenta com tecnologia de IA para limpar painéis de mangá. ColorButton - + Select Color Selecionar cor @@ -667,7 +667,7 @@ Uma ferramenta com tecnologia de IA para limpar painéis de mangá.Sistema não suportado - + This system is not supported. Esse sistema não é compatível. @@ -682,12 +682,12 @@ Uma ferramenta com tecnologia de IA para limpar painéis de mangá.Desinstalação do Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. A extensão será instalada em {path}. Isso não requer'permissões de administrador. - + The extension file will be deleted. This doesn't require root permissions. O arquivo de extensão será excluído. Isso não requer'permissões de administrador. @@ -702,17 +702,17 @@ Uma ferramenta com tecnologia de IA para limpar painéis de mangá.Desinstalar para Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. A extensão será instalada no registro do Windows, o que requer permissões de administrador. - + The extension will be uninstalled from the Windows registry.This requires admin permissions. A extensão será desinstalada do registro do Windows, o que requer permissões de administrador. - + Dolphin extension installed successfully. Extensão do Dolphin instalada com sucesso. @@ -722,7 +722,7 @@ Uma ferramenta com tecnologia de IA para limpar painéis de mangá.Instalação bem-sucedida - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. A extensão do Dolphin foi instalada com sucesso. @@ -802,12 +802,12 @@ Por padrão, ele usa o perfil incorporado. Entretanto, você pode alterar essa c FileTable - + Original size in pixels (width × height) Tamanho original em pixels (largura × altura) - + Processing size in pixels (width × height), scale factor Tamanho do processamento em pixels (largura × altura), fator de escala @@ -833,7 +833,7 @@ Por padrão, ele usa o perfil incorporado. Entretanto, você pode alterar essa c Arquivos TIFF não suportados - + The following 5-channel TIFF files are not supported: Os seguintes arquivos TIFF de 5 canais não são compatíveis: @@ -860,27 +860,27 @@ Por padrão, ele usa o perfil incorporado. Entretanto, você pode alterar essa c dividir - + Number of boxes removed by the OCR model / total boxes Número de balões removidos pelo modelo OCR / total de balões - + Number of boxes that failed to generate a mask / total boxes Número de balões que não conseguiram gerar uma máscara / total de balões - + Number of boxes that were perfectly masked / total boxes Número de balões que foram perfeitamente mascarados / total de balões - + Number of boxes that were denoised / total boxes Número de balões que foram tratados (denoise)/ total de balões - + Number of boxes that were inpainted Número de balões que passaram por Inpaint @@ -1028,14 +1028,14 @@ Por padrão, ele usa o perfil incorporado. Entretanto, você pode alterar essa c Solicitação impossível - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. A redução de ruído está desativada no perfil atual, essa saída pode't ser gerada. Ative a redução de ruído nas configurações do perfil e tente novamente. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting está desativada no perfil atual, essa saída não pode ser gerada.' @@ -1057,7 +1057,7 @@ Ative a pintura nas configurações do perfil e tente novamente. Falha na exportação - + Failed to export image: Falha ao exportar a imagem: @@ -1100,7 +1100,7 @@ Ative a pintura nas configurações do perfil e tente novamente. resultado - + results resultados @@ -1115,17 +1115,17 @@ Ative a pintura nas configurações do perfil e tente novamente. nova imagem - + new images novas imagens - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Correspondência entre {num_images} {plural_images} e os resultados do OCR. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} OCR órfão {plural_result} será perdido. @@ -1224,12 +1224,12 @@ Ative a pintura nas configurações do perfil e tente novamente. Observação: O nome "{name}" estava oculto - + Corrupted log session Sessão de registro corrompida - + Current session Sessão atual @@ -1244,7 +1244,7 @@ Ative a pintura nas configurações do perfil e tente novamente. Erro - + Errors Erros @@ -1254,7 +1254,7 @@ Ative a pintura nas configurações do perfil e tente novamente. Crítico - + Criticals Críticos @@ -1270,49 +1270,49 @@ Ative a pintura nas configurações do perfil e tente novamente. MainWindow - + Open Config Statusbar button Abrir configuração - + Open Log Statusbar button Abrir registro - + CUDA Enabled CUDA Habilitado - + Select Output Directory Selecione o diretório de saída - + Aborting... Cancelando... - + Loading OCR model... Carregando o modelo de OCR... - + Loaded OCR model. Modelo de OCR carregado. - + Error Erro - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1321,7 +1321,7 @@ Continue anyway? Continuar mesmo assim? - + Delete Models Excluir modelos @@ -1331,23 +1331,23 @@ Continuar mesmo assim? Tem certeza de que deseja excluir os modelos de aprendizado de máquina? Isso impossibilitará a limpeza e o OCR até que eles sejam baixados novamente. - + Encountered error: Erro encontrado: - + MPS Enabled MPS ativado - - + + Inpainting Model Update Atualização do Modelo de Inpaining - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1361,148 +1361,148 @@ Mudar para o novo modelo? Modelo antigo mantido. Para atualizar, selecione "Ajudar" e depois "Excluir Modelos de Aprendizagem de Máquina" no menu. - + Layout Reset Redefinição de layout - + The window layout has been reset. The default layout will be used next time. O layout da janela foi redefinido. O layout padrão será usado na próxima vez. - + Multiple Instances Várias instâncias - + Found process ID in lock file: ID de processo encontrado no arquivo de bloqueio: - + Failed to delete models. Falha ao deletar o modelos - + Delete Failed Falha na Exclusão - + Models Deleted Modelos excluídos - + The models were deleted. Would you like to download them again? Os modelos foram excluídos. Gostaria de baixá-los novamente? - + Not Supported Não suportado - + Unsupported system or package format Sistema ou formato de pacote não suportado - + Shutdown Desligar - + After processing, the system will shut down. Após o processamento, o sistema será desligado. - + You have review options enabled, these will need to be manually closed before the action can start. Se as opções de revisão estiverem ativadas, elas precisarão ser fechadas manualmente para que a ação possa ser iniciada. - + You have disabled writing output. Você desativou a saída de gravação. - + Shutting down system... Desligando o sistema... - + Running post action "{action}"... Executando a ação "{action}"... - + Restart Required Reinicialização necessária - + The language has been changed. Please restart the application for the changes to take effect. O idioma foi alterado. Reinicie o aplicativo para que as alterações tenham efeito. - - + + System Language Idioma do sistema - + Import Profile Importar um Perfil - + Profile Files (*.conf) Arquivos de perfil (*.conf) - + Profile Imported Perfil Importado - + Import Error Erro de importação - + Failed to Delete Falha ao excluir - + The default profile cannot be deleted. O perfil padrão não pode ser excluído. - + Delete Profile Excluir perfil - + Are you sure you want to delete the profile {profile_name}? Tem certeza de que deseja excluir o perfil {profile_name}? - + Delete Error Erro ao excluir - + Failed to delete the profile. Falha ao excluir o perfil. @@ -1514,217 +1514,217 @@ Switching profiles will discard changes to the current profile. A troca de perfis descartará as alterações feitas no perfil atual. - + Unsaved changes Alterações não salvas - - + + Load Error Erro ao carregar - - + + Failed to load profile: Falha ao carregar o perfil: - + Tesseract OCR is not installed or not found O Tesseract OCR não está instalado ou não foi encontrado - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Não é possível'usar o Tesseract para executar o OCR. Revertendo para manga-ocr. Consulte as instruções para instalar o Tesseract corretamente <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> ou continue usando o modelo padrão.</html> - + Unsupported Language Idioma não suportado - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. O idioma '{lang_name}' não é compatível com nenhum dos mecanismos de OCR atuais. Consulte a documentação on-line para saber como adicionar suporte a mais idiomas. - + Profile Created Perfil criado - + Create Error Erro ao criar - + Failed to save profile to {profile_path} Falha ao salvar o perfil em {profile_path} - - + + Save Error Erro ao salvar - + Failed to save profile. Falha ao salvar o perfil. - + Profile saved to {profile_path} Perfil salvo em {profile_path} - + Failed to save config. Falha ao salvar a configuração. - + Failed to save the new profile to the configuration file. Falha ao salvar o novo perfil no arquivo de configuração. - - + + No Files Sem arquivos - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. Não há arquivos para processar. Você pode adicionar arquivos arrastando-os e soltando-os no meio da janela ou por meio da barra de menus: Arquivos -> Adicionar arquivos ou Adicionar pasta. - + No Outputs Sem saídas - + No outputs were requested. Please select at least one output before cleaning. Nenhuma saída foi solicitada. Selecione pelo menos uma saída antes de limpar. - - + + cleaned limpo - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. Não há arquivos para processar. Para editar um arquivo de saída de OCR antigo, você deve primeiro carregar (uma ou mais) as imagens às quais ele corresponde. - + Open OCR Output File Abrir arquivo de saída de OCR - + OCR Output Files (*.txt *.csv) Arquivos de saída de OCR (*.txt *.csv) - + Parse Error Erro de análise - + Failed to parse the OCR output file. Falha ao analisar o arquivo de saída de OCR. - + File Exists O arquivo já existe - + The file '{output_path}' already exists. Overwrite? O arquivo '{output_path}' já existe. Substituir? - + Export Images Exportar imagens - + Would you like to export the cleaned images? Você gostaria de exportar as imagens limpas? - + Export OCR Results Exportar resultados de OCR - + After processing, the "{action}" action will be executed. Após o processamento, a ação "{action}" será executada. - - + + Post action canceled. Pós-ação cancelada. - + Un-Applied Changes Alterações não aplicadas - + You have un-applied changes to the profile. Continue anyway? Você não aplicou alterações no perfil. Continuar mesmo assim? - + Would you like to export the OCR results? Você gostaria de exportar os resultados do OCR? - + Processing Finished Processamento concluído - + Finished processing all files. Processamento de todos os arquivos, completo. - + Processing Aborted Processamento abortado - + Processing aborted. Processamento cancelado. - + Processing Error Erro de processamento - + Encountered an error while processing files. Encontrou um erro ao processar arquivos. @@ -1735,7 +1735,7 @@ Consulte a documentação on-line para saber como adicionar suporte a mais idiom - + Profile Perfil @@ -1750,360 +1750,365 @@ Consulte a documentação on-line para saber como adicionar suporte a mais idiom Aplicar - - + + Save Salvar - + Reset All Redefinir tudo - + Drag and Drop Images or Folders Here Arraste e solte imagens ou pastas aqui - + Images Imagens - - + + File Arquivo - + Size Tamanho - + Processing Size The size that the picture is shrunk down to for processing. Tamanho do processamento - + File Size Tamanho do arquivo - + Color Mode Like RGB, grayscale etc. Modo de cor - + Analytics Can also call this statistics. Análises - + Process Processo - + Cleaning Limpando - + OCR OCR - + Edit Existing Output Editar saída existente - + Abort Interromper - + Start Iniciar - + Output Saída - + Cleaned Image Imagem limpa - + Mask Máscara - + Isolated Text Texto isolado - - + + Review Output Revisar Saída - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>Quando marcada, uma janela de revisão será aberta após a conclusão do processo. Os resultados também podem ser visualizados previamente em cada imagem's visão detalhada, antes de continuar a salvá-las.</p></body></html> - + Save Output Salvar saída - + Output Directory: Diretório de saída: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. Você pode usar um caminho relativo para criar uma subpasta no local original da imagem's ou usar um caminho absoluto. - + Plain Text Texto sem Formatação - + When checked, a review window will open upon process completion. Quando marcada, uma janela de revisão será aberta após a conclusão do processo. - + Output File: Arquivo de saída: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Insira um caminho absoluto ou relativo com um nome de arquivo. A saída de todas as imagens é gravada no mesmo arquivo.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. texto_detectado.txt - + Cancel Action Cancelar ação - + Close Fechar - + Generating: This is present progressive, as in "[currently] Generating:" Gerando - + Current Step: Etapa atual: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Imagens processadas: - + Set Default Definir padrão - + Settings Configurações - + Theme As in color theme Tema - + Language Idioma - + Help Ajuda - + Add Files... Adicionar arquivos... - + Remove All Files Remover todos os arquivos - + Remove File Remover arquivo - + Help Translate Panel Cleaner Ajude a traduzir o Panel Cleaner - + Report an Issue... Relatar um problema... - + Simulate Exception Simular exceção - + Add File Browser Extension... Adicionar extensão do navegador de arquivos... - + Show Supported OCR Languages Mostrar idiomas de OCR suportados - - + + Reset Window Layout Preferences Redefinir preferências de layout de janela - + Show Out Of Memory Warnings Mostrar avisos de falta de memória - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Ações pós-execução - + Configure Post-Run Actions... Configurar Ações pós-execução... - + New Novo - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>Quando marcada, as saídas são salvas no disco. Caso contrário, você só poderá visualizá-las na exibição de detalhes da imagem ou na Revisão de saída, se ativada. </p><p>Se o perfil permanecer inalterado após a limpeza sem essa opção ativada, você poderá exportá-lo rapidamente executando o limpador novamente com essa opção ativada.</p></body></html> - + CSV Spreadsheet Planilha CSV - + Delete Excluir - + Import... Importar... - + Online Documentation Documentação on-line - + About Panel Cleaner Sobre o Panel Cleaner - + View License Exibir licença - + Save as... Salvar como... - + System Sistema - + Dark Escuro - + Light Luz - + Add Folder... Adicionar pasta... - + Delete Machine Learning Models Excluir Modelos de Aprendizagem de Máquina (LLM) - + Download Machine Learning Models Baixar Modelos de Aprendizagem de Máquina (LLM) - + Donate Faça uma Doação @@ -2135,19 +2140,19 @@ Consulte a documentação on-line para saber como adicionar suporte a mais idiom Modelo OCR: - + Already downloaded Já baixado - + Text Detector model (CUDA) Modelo de detector de texto (CUDA) - + Text Detector model (CPU) Modelo de detector de texto (CPU) @@ -2166,7 +2171,7 @@ Consulte a documentação on-line para saber como adicionar suporte a mais idiom Erros de download de OCR - + ETA estimated time of completion Tempo Estimado @@ -2744,7 +2749,7 @@ Consulte a documentação on-line para saber como adicionar suporte a mais idiom O nome "{shutdown_command}" é reservado. - + The Flatpak sandbox will likely prevent actions from working. O sandbox do Flatpak provavelmente impedirá que as ações funcionem. @@ -2825,73 +2830,73 @@ Consulte a documentação on-line para saber como adicionar suporte a mais idiom Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input A imagem original com dimensionamento aplicado (se necessário). - + Input Step name in the image details view Entrada - + Input Output name in the image details view Entrada - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask A máscara bruta gerada pela IA. - + Text Detection Step name in the image details view Detecção de texto - + AI Mask Output name in the image details view Máscara de IA - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes Os dados de balões não filtrados gerados pela IA. - + Raw Boxes Output name in the image details view Balões na Raw - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes Contornos dos balões que a IA encontrou. - + Preprocessor Step name in the image details view Pré-processador - + Initial Boxes Output name in the image details view Balões iniciais - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2899,181 +2904,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Verde: balões iniciais. Vermelho: balões estendidos. Roxo: balões mesclados (finais). Azul: balões de referência para redução de ruído. - + Final Boxes Output name in the image details view Balões finais - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask A máscara dos balões mesclados. - + Masker Step name in the image details view Mascarador - + Box Mask Output name in the image details view Máscara de quadrado - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask Máscara de detecção preliminar de texto, com as áreas fora da máscara da caixa excluídas. - + Cut Mask Output name in the image details view Máscara de corte - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers As diferentes etapas de expansão em torno da máscara de corte são exibidas em cores diferentes. - + Mask Layers Output name in the image details view Camadas de máscara - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality O desvio padrão (σ) e a espessura do contorno (em pixels) de cada melhor máscara selecionada, se houver. Valores menores de σ indicam melhor qualidade, variando de perfeito (roxo) a falho (vermelho). - + Fitment Quality Output name in the image details view Qualidade do ajuste - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay A imagem original com a máscara final aplicada por cima, em cores. - + Mask Overlay Output name in the image details view Sobreposição de máscara - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask A coleção de máscaras para cada balão que melhor se adapta. - + Final Mask Output name in the image details view Máscara final - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text A camada de texto isolada da imagem de entrada. - + Isolated Text Output name in the image details view Texto isolado - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output A imagem original com a máscara final aplicada. - + Masked Output Output name in the image details view Saída com a máscara - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask Máscaras que passaram por remoção de ruído, destinadas a serem sobrepostas na máscara final durante a exportação. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Máscara de Tratamento (Denoise) - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output A imagem original com redução de ruído aplicado na máscara. - + Denoised Output Output name in the image details view Imagem Tratada (Denoised) - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask As áreas preenchidas (inpaint) ao redor do texto que não foi bem limpo, se existirem. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Máscara de Inpaiting - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output A imagem original: limpa, tratada (se ativado) e inpainted. - + Inpainted Output Output name in the image details view Resultado do Inpaint - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3082,1141 +3087,1141 @@ Verde: balões iniciais. Vermelho: balões estendidos. Roxo: balões mesclados ( Profile - + General Profile section title Geral - + About this profile: Profile comment in General Sobre este perfil: - + Notes Profile entry in General Anotações - + Preferred file type to save the cleaned image as. Profile comment in General Tipo de arquivo preferido para salvar a imagem limpa. - + Preferred File Type Profile entry in General Tipo de arquivo preferido - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Tipo de arquivo preferido para salvar a máscara. Somente formatos de imagem que permitem transparência são suportados. - + Preferred Mask File Type Profile entry in General Tipo de arquivo de máscara preferido - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General A seguir estão os limites inferior e superior para a altura da imagem original. A imagem é reduzida apenas para se ajustar a esse intervalo, dando preferência a fatores de número inteiro para minimizar o impacto na qualidade da imagem. Imagens menores que qualquer um dos limites permanecerão inalteradas. Você pode desativar esse recurso definindo um ou ambos os valores como menores ou iguais a 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General Isso é útil para acelerar significativamente o processamento de imagens grandes. Além disso, como outras opções que dependem de dimensões de pixel dependem do tamanho, isso ajudará a normalizar os resultados em diferentes tamanhos de imagem. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General A imagem será diminuida, processada e, em seguida, apenas a máscara será redimensionada de volta. Isso significa que a saída limpa ainda usará a imagem original, sem redimensionamento, para evitar qualquer perda de qualidade. Apenas a altura da imagem é usada para determinar o fator de escala, preservando a proporção e ignorando a largura individual da imagem, de forma que o fator permaneça consistente caso uma das páginas seja uma página dupla. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General Por exemplo, para um limite inferior de 1000 e um limite superior de 2000, uma imagem com tamanho de 5000x7000 pixels (largura, altura) será redimensionada por um fator de 4, passando a ter o tamanho de 1250x1750 pixels durante o processamento. - + Input Height Lower Target Profile entry in General Limite Inferior de Altura - + Input Height Upper Target Profile entry in General Limite Superior de Altura - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Dividir páginas longas em páginas individuais. Se ativada, em vez de comprimir toda a página para ajustá-la à altura preferida, a página será dividida em páginas individuais, cada uma ajustada à altura preferida mais/menos o dobro da margem de tolerância de corte (se o segmento estiver no meio da página). - + Split Long Strips Profile entry in General Cortar Páginas Longas - + Preferred height to split long strips at. Profile comment in General Altura preferida para cortar páginas longas - + Preferred Split Height Profile entry in General Altura de corte preferida - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Margem de tolerância para corte de páginas longas. Essa é a diferença máxima entre a altura de corte preferido e a altura real da página. Um algoritmo determina o melhor ponto de corte dentro dessa margem, em um esforço para evitar um corte no meio de um painel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Proporção de aspecto a ser usada para cortar páginas longas. Essa é a relação entre a largura e a altura da imagem. Se a proporção da imagem'for menor que esse valor, ela será considerada uma página longa.<br> Exemplo: 0,3 significa que a largura da imagem é 0,3 vezes a altura. - + Long Strip Aspect Ratio Profile entry in General Proporção de Página Longa - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Mesclar páginas longas em uma única imagem. Se ativado, as páginas individuais criadas a partir de uma faixa longa serão mescladas em uma única imagem na exportação. - + Merge After Split Profile entry in General Mesclar após cortar - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Número máximo de threads a serem usados para exportar imagens. Você pode deixar esse valor não especificado para usar todos os threads disponíveis. Diminua esse valor se tiver problemas de memória, que aparecerão como travamentos aleatórios. - + Max Threads Export Profile entry in General Exportação máxima de threads - + Text Detector Profile section title Detecção de texto - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Caminho para o modelo de detecção de texto; deixe em branco para usar o modelo incorporado. Recomenda-se substituir isso somente se a versão baixada automaticamente for mais antiga do que a versão mais recente. O caminho deve apontar diretamente para o arquivo comictextdetector.pt (CUDA) ou comictextdetector.pt.onnx (CPU). Você pode fazer download de versões mais antigas do modelo <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">aqui.</a> - + Model Path Profile entry in Text Detector Caminho do modelo - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Número de modelos a serem executados em paralelo. Isso é útil se você tiver RAM (ou VRAM com CUDA) suficiente para executar vários modelos ao mesmo tempo. Isso, é claro, aumentará a velocidade do processo, mas também poderá travar o computador se você superestimar o hardware. Recomendo usar 1 modelo para cada 2 GB de memória disponível ou 1 modelo para cada 50 imagens. Observação: Isso é ignorado se o processamento for inferior a 50 arquivos, pois a sobrecarga de iniciar vários modelos não vale a pena abaixo disso. Aviso: Isso pode causar instabilidade no programa; use por sua própria conta e risco. <br>NÃO relate problemas sobre essa configuração, pois ela'depende totalmente do hardware! - + Concurrent Models Profile entry in Text Detector Modelos simultâneos - + Preprocessor Profile section title Pré-processador - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Os tamanhos das caixas são fornecidos no número total de pixels, portanto, uma caixa de 200x200 pixels tem um tamanho de 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Tamanho mínimo de qualquer caixa para mantê-la. - + Box Min Size Profile entry in Preprocessor Tamanho mínimo da caixa - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Tamanho mínimo de uma caixa com "linguagem" desconhecida para mantê-la. Essa linguagem é normalmente atribuída a logotipos e outros elementos de texto com comportamento inadequado. - + Suspicious Box Min Size Profile entry in Preprocessor Tamanho Mínimo da Caixa Suspeita - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor A sobreposição mínima entre duas caixas para mesclá-las. Essa porcentagem (0-100) significa quanto da caixa menor deve estar dentro da caixa maior para ser mesclada. Um valor mais alto exigirá uma sobreposição maior para mesclar as caixas. - + Box Overlap Threshold Profile entry in Preprocessor Limite de sobreposição de caixa - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Usar OCR para detectar caixas que'não valem a pena limpar, como as que contêm apenas números ou símbolos? - + OCR Enabled Profile entry in Preprocessor OCR Habilitado: - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Usar o Tesseract para executar tarefas de OCR.<br> Se marcada, o Tesseract OCR pode ser usado para extração de texto, se disponível.<br> Se desmarcada, o modelo de OCR incorporado (manga-ocr) é sempre usado, o que é mais adequado para texto japonês vertical. - + OCR Use Tesseract Profile entry in Preprocessor Usar Tesseract para OCR - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor O idioma a ser usado nas tarefas de OCR. O detector de texto pode detectar japonês e inglês; para qualquer outro idioma, selecione um idioma explicitamente aqui. Detectar por caixa retém o que o detector de texto detectou, enquanto que por página atribuirá o idioma mais proeminente à página inteira.<br> Observação: Apenas o japonês é suportado imediatamente, todo o resto requer que o Tesseract seja ativado e que seus pacotes de idiomas associados sejam instalados. - + OCR Language Profile entry in Preprocessor Idioma para OCR - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Especifica o mecanismo a ser usado para executar o OCR.<br> - auto: Seleciona automaticamente o mecanismo de OCR com base no idioma detectado de cada bloco de texto na imagem. Usa o Manga Ocr para texto em japonês e o Tesseract para texto em inglês ou desconhecido.<br> - mangaocr: Força o Panel Cleaner a usar o modelo manga-ocr incorporado para todas as tarefas de reconhecimento de texto. Mais adequado para texto vertical em japonês.<br> - tesseract: Força o Panel Cleaner a usar o Tesseract OCR para todas as tarefas de reconhecimento de texto. Essa é uma opção versátil que oferece suporte ao inglês e a vários outros idiomas. - + OCR Engine Profile entry in Preprocessor Mecanismo de OCR - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Define a ordem de leitura para processamento e classificação de caixas de texto em toda a página, não em blocos de texto individuais. Essa configuração global influencia a forma como as caixas de texto são ordenadas e apresentadas para processamento posterior.<br> - auto: Detecta a ordem de leitura com base no idioma detectado de cada bloco de texto na página.<br> - mangá: Ordem da direita para a esquerda, de cima para baixo. Adequado para mangás japoneses.<br> - quadrinhos: Da esquerda para a direita, ordem de cima para baixo. Adequado para quadrinhos e textos ocidentais.<br> Escolha com base no layout predominante do seu conteúdo. - + Reading Order Profile entry in Preprocessor Ordem de leitura - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Tamanho máximo de uma caixa para executar o OCR. Caixas inúteis geralmente são pequenas, e o OCR é lento, portanto, use isso como um limite. - + OCR Max Size Profile entry in Preprocessor Tamanho máximo do OCR - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Padrão Regex para comparar com os resultados do OCR. Tudo o que corresponder a esse padrão será descartado.<br> Observação: o modelo MangaOCR retorna caracteres de largura total, portanto, esse padrão deve corresponder a eles. - + OCR Blacklist Pattern Profile entry in Preprocessor Lista negra de padrões para o OCR - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor O modelo MangaOCR só pode lidar com texto em japonês, portanto, quando a opção "estrita" estiver ativada, ele descartará as caixas em que o Detector de texto não estiver'confiante de que são japonesas. Às vezes, números ou outros símbolos diminuem a confiança, fazendo com que o idioma detectado seja desconhecido. Se strict estiver desativado, eles não serão descartados. Qualquer coisa que seja reconhecida com confiança como um idioma diferente será descartada independentemente dessa configuração.<br> Observação: essa configuração só é relevante quando ocr_language está definido para detectar por caixa ou página. - + OCR Strict Language Profile entry in Preprocessor Idioma estrito do OCR - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Espaçamento para adicionar ao lado de uma caixa. Isso é adicionado às caixas iniciais criadas pelo detector de texto AI. Essas caixas são visualizadas em verde. - + Box Padding Initial Profile entry in Preprocessor Espaçamento de Caixa Inicial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Espaçamento para adicionar ao lado direito de uma caixa. Essa extensão ajuda a cobrir o texto em rubi que flutua à direita do texto vertical. - + Box Right Padding Initial Profile entry in Preprocessor Espaçamento Inicial à Direita - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Espaçamento para adicionar a cada lado de uma caixa. Isto é adicionado a um conjunto estendido de caixas, usado para cortar falsos positivos pelo detector de texto AI's máscara. Estas caixas são visualizadas em roxo. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Espaçamento Direito da Caixa Estendido - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Espaçamento para adicionar a cada lado de uma caixa. Isso é adicionado às caixas de referência usadas para a amostra da imagem original enquanto analisa qual máscara usar. Estas caixas são visualizadas em azul. - + Box Reference Padding Profile entry in Preprocessor Espaçamento de Referência - + Masker Profile section title Mascarador - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Número máximo de threads a serem usados para a geração de máscaras. Você pode deixar esse valor não especificado para usar todos os threads disponíveis. Diminua esse valor se tiver problemas de memória, que aparecerão como travamentos aleatórios. - + Max Threads Profile entry in Masker Máximo de threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Número de pixels para aumentar a máscara em cada etapa. Isso aumenta o contorno da máscara, portanto, valores menores serão mais precisos, porém mais lentos. - + Mask Growth Step Pixels Profile entry in Masker Pixels para crescimento de máscara - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Número de etapas para aumentar a máscara. Um número maior criará máscaras cada vez maiores, limitadas, em última análise, pelo tamanho da caixa de referência. - + Mask Growth Steps Profile entry in Masker Etapas de crescimento da máscara - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Espessura mínima de uma máscara.'Essa é a primeira etapa de crescimento da máscara, e as demais seguirão a etapa de crescimento da máscara em pixels. Dessa forma, você pode ter um tamanho de etapa pequeno para obter precisão, mas ainda assim evitar máscaras muito finas que possam ser usadas para limpar o texto apenas cercado por um contorno, quando Inpainting seria'a melhor opção. - + Min Mask Thickness Profile entry in Masker Espessura mínima da máscara - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter O desvio padrão mínimo de cores ao redor da borda de uma determinada máscara para executar a pintura na região ao redor da máscara. Se esse valor corresponder ao desvio máximo para máscaras, somente os balões com falha serão pintados, tornando as duas configurações a seguir irrelevantes. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine os resultados em um único arquivo de projeto como camadas. Formatos atualmente suportados: Photoshop PSD.<br> - nenhum: Cada imagem e máscara são salvas como arquivos básicos.<br> - psd-per-image: As imagens e as máscaras são salvas juntas em um arquivo PSD por imagem de entrada.<br> - psd-bulk: Todas as imagens e máscaras são salvas juntas em um único arquivo PSD, agrupadas por imagem de entrada. - + Layered Export Profile entry in General Exportação em camadas - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Permitir máscaras coloridas? Quando habilitado, máscaras usaram qualquer cor, não apenas branco, preto ou cinza. - + Allow Colored Masks Profile entry in Masker Permitir máscaras coloridas - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Limite máximo para que um pixel seja considerado esbranquiçado. A cor mediana ao longo da borda de uma máscara pode não ser branco puro, portanto, para evitar tons levemente esbranquiçados em um fundo branco puro, qualquer coisa mais clara do que esse valor de limite será arredondada para branco puro. Intervalo de valores: preto (0) a branco puro (255). - + Off White Max Threshold Profile entry in Masker Limite máximo tons de branco - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker O desvio padrão de uma máscara representa a variação de cor ao longo da borda da máscara. Para isso, apenas a linha única de pixels ao longo da borda externa de uma máscara é amostrada da imagem original.'Um desvio baixo significa que a máscara fica em uma cor sólida, o que significa que ela não cruza nenhum texto ou outros objetos. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Melhoria mínima no desvio padrão da máscara para continuar a aumentá-la. Definir um valor mais alto aqui requer uma melhoria maior para considerar uma máscara maior, para dar preferência a máscaras menores. - + Mask Improvement Threshold Profile entry in Masker Limite de aprimoramento da máscara - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Se deve ser usado o algoritmo de seleção rápida de máscara. Quando verdadeiro, o algoritmo de seleção de máscaras escolherá a primeira máscara perfeita, se uma for encontrada com antecedência. Isso é mais rápido, mas pode não encontrar a melhor máscara, se uma um pouco maior for melhor. - + Mask Selection Fast Profile entry in Masker Seleção rápida de máscaras - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Número máximo de threads a serem usados para a redução de ruído. Você pode deixar esse valor não especificado para usar todos os threads disponíveis. Diminua esse valor se tiver problemas de memória, que aparecerão como travamentos aleatórios. - + Max Threads Profile entry in Denoiser Máximo de threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter RECURSO EXPERIMENTAL: Se você encontrar configurações padrão melhores, abra um problema no github para compartilhar as melhorias com todos. Observe que o inpainting é't ativado por padrão. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting é quando o aprendizado de máquina é usado para substituir o conteúdo de uma imagem com base em seus arredores. Para máscaras que'não puderam ser bem limpas (ou não puderam ser limpas de forma alguma), o inpainting pode ser usado. Para diferenciar isso da redução de ruído, o inpainting destina-se a máscaras significativamente piores que tiveram um ajuste apertado. Quaisquer máscaras que tenham sido eliminadas não'serão afetadas. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Como essa etapa pode fornecer resultados ruins em alguns casos, ela pode ser desativada aqui. Desmarque a opção para desativar a pintura. - + Inpainting Enabled Profile entry in Inpainter Inpainting ativado - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Mínimo Desvio padrão - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter O raio máximo de uma máscara para executar a pintura. As máscaras maiores do que isso serão deixadas como estão, pois se a margem for tão grande, é provável que a máscara já esteja boa o suficiente. - + Inpainting Max Mask Radius Profile entry in Inpainter Raio máximo da máscara de inpaint - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter O raio mínimo em torno de uma máscara para inpaint. Isso é adicionado ao tamanho ideal da máscara para garantir que a pintura cubra toda a máscara. - + Min Inpainting Radius Profile entry in Inpainter Raio mínimo de inpaint - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter Para máscaras que se mostraram muito mais difíceis de limpar, o que significa que tinham um alto desvio padrão, aumente o raio da pintura para cobrir mais da máscara. Essa margem adicional é adicionada ao raio mínimo de pintura e é calculada como: multiplicador do raio de pintura vezes o desvio padrão da máscara - + Inpainting Radius Multiplier Profile entry in Inpainter Multiplicador de raio de inpainting - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter O raio máximo em torno de uma máscara para inpaint. Isso limita o tamanho que a pintura pode atingir, para evitar que ela cubra muito a imagem, se um multiplicador de raio grande for usado. - + Max Inpainting Radius Profile entry in Inpainter Raio máximo de inpaint - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter Após o inpaint, corte o resultado da imagem original para evitar que o inpaint afete o restante da imagem. Isso garante que a imagem original seja preservada o máximo possível. Esse raio é adicionado ao redor do raio final do inpaint, devido ao fato de o modelo de inpaint modificar alguns pixels fora de sua região dedicada. - + Inpainting Isolation Radius Profile entry in Inpainter Raio de isolamento de inpainting - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Suaviza as bordas da máscara de inpaint com essa quantidade de pixels para mesclar suavemente as partes afetadas com inpaint com o restante da imagem.<br> Se você vir contornos fracos após o inpainting, aumente o valor mínimo de inpainting e defina este valor como a metade. - + Inpainting Fade Radius Profile entry in Inpainter Raio de suavização de inpainting - + Detect per box Language option for OCR Detectar por caixa - + Detect per page Language option for OCR Detectar por página - + Japanese Language option for OCR Japonês - + English Language option for OCR Inglês - + Korean Language option for OCR Coreano - + Korean (vertical) Language option for OCR Coreano (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinês Tradicional - + Albanian Language option for OCR Albanês - + Arabic Language option for OCR Árabe - + Azerbaijani Language option for OCR Azerbaijão - + Azerbaijani - Cyrilic Language option for OCR Azerbaijão - cirílico - + Bengali Language option for OCR Bengalês - + Bulgarian Language option for OCR Búlgaro - + Burmese Language option for OCR Birmanês - + Catalan; Valencian Language option for OCR Catalão; Valenciano - + Croatian Language option for OCR Croata - + Czech Language option for OCR Tcheco - + Danish Language option for OCR Dinamarquês - + Dutch; Flemish Language option for OCR Holandês; flamengo - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estoniano - + Finnish Filipino Language option for OCR Finlandês - + French Language option for OCR Francês - + Georgian Language option for OCR Georgiano - + German Language option for OCR Alemão - + Greek Language option for OCR Grego - + Hebrew Language option for OCR Hebraico - + Hindi Language option for OCR Hindi (हिन्दी) - + Hungarian Language option for OCR Húngaro - + Indonesian Language option for OCR Indonésio - + Italian Language option for OCR Italiano - + Kazakh Language option for OCR Cazaquistão - + Latin Language option for OCR Latim - + Lithuanian Language option for OCR Lituanês - + Luxembourgish Language option for OCR Luxemburguês - + Malay Language option for OCR Malaio - + Mongolian Language option for OCR Mongol - + Nepali Language option for OCR Nepalês - + Norwegian Language option for OCR Norueguês - + Persian Language option for OCR Persa (فارسی) - + Polish Language option for OCR Polonês - + Portuguese Language option for OCR Português - + Romanian; Moldavian Language option for OCR Romeno; moldavo - + Russian Language option for OCR Russo - + Serbian Language option for OCR Sérvio - + Serbian - Latin Language option for OCR Sérvio - Latim - + Slovak Language option for OCR Eslovaco - + Slovenian Language option for OCR Esloveno - + Spanish; Castilian Language option for OCR Espanhol; Castelhano - + Swedish Language option for OCR Sueco - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tâmil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Tailandês - + Turkish Language option for OCR Turco - + Ukrainian Language option for OCR Ucraniano - + Vietnamese Language option for OCR Vietnamita - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker O desvio padrão máximo de uma máscara a ser considerado. Um valor alto aqui significa uma tolerância maior para a interseção da máscara com o texto ou outros objetos, o que'não é uma boa máscara, pois exigirá pintura de qualquer forma. Definir esse valor como 0 só permitirá máscaras perfeitas, o que é recomendado para imagens de altíssima resolução. - + Mask Max Standard Deviation Profile entry in Masker Máscara Desvio padrão máximo - + Color to use for the debug mask. Profile comment in Masker Cor a ser usada para a máscara de depuração. - + Debug Mask Color Profile entry in Masker Cor da máscara de depuração - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser Quando um texto é difícil de mascarar, a etapa de limpeza provavelmente teve que escolher uma máscara pequena, o que deixa muitos artefatos de jpeg para trás, se eles estavam ao redor do texto. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser 'Isso significa que a cor dos pixels ao redor da borda de uma máscara não é uniforme, o que é quantificado como um desvio padrão. O denoiser pode direcionar as máscaras com um desvio padrão mínimo e fazer o denoise da área ao redor da máscara. Isso preserva os detalhes no restante da imagem, mas remove os artefatos ao redor do local onde o texto costumava estar. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Como essa etapa é opcional e pode até ser supérflua para imagens de alta resolução que não'sofrem com artefatos jpeg, ela pode ser desativada aqui. Defina como Falso para desativar a redução de ruído. - + Denoising Enabled Profile entry in Denoiser Denoising ativado - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser O desvio padrão mínimo de cores ao redor da borda de uma determinada máscara para realizar a redução de ruído na região ao redor da máscara. - + Noise Min Standard Deviation Profile entry in Denoiser Ruído Desvio padrão mínimo - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Observação: Se Inpainting estiver ativado, o desvio padrão mínimo da pintura funcionará como um máximo para isso, após o qual essa máscara será entregue ao Inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser A espessura de um contorno a ser eliminado em torno de uma máscara. - + Noise Outline Size Profile entry in Denoiser Tamanho do contorno do ruído - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Suavize as bordas da máscara com redução de ruído nessa quantidade para mesclar suavemente as partes com redução de ruído com o restante da imagem. - + Noise Fade Radius Profile entry in Denoiser Raio de suavização de ruído - + Set to True to support colored images. Profile comment in Denoiser Defina como verdadeiro para oferecer suporte a imagens coloridas. - + Colored Images Profile entry in Denoiser Imagens coloridas - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser Qual a intensidade da redução de ruído da imagem. Valores mais altos significam mais redução de ruído. - + Filter Strength Profile entry in Denoiser Força do filtro - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser A intensidade da redução de ruído dos canais de cores, em oposição à luminosidade. Valores mais altos significam mais redução de ruído. - + Color Filter Strength Profile entry in Denoiser Força do filtro de cor - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Tamanho em pixels do patch do modelo que é usado para calcular pesos. Deve ser ímpar. - + Template Window Size Profile entry in Denoiser Tamanho da janela do modelo - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Tamanho em pixels da janela que é usada para calcular a média ponderada de um determinado pixel. Deve ser ímpar. - + Search Window Size Profile entry in Denoiser Tamanho da janela de pesquisa @@ -4243,7 +4248,7 @@ Verde: balões iniciais. Vermelho: balões estendidos. Roxo: balões mesclados ( px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. Todos os núcleos @@ -4660,14 +4665,14 @@ Verde: balões iniciais. Vermelho: balões estendidos. Roxo: balões mesclados ( ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>O caminho do modelo {value} do Detector de Texto não existe, revertendo para o padrão. Você pode baixar o modelo manualmente de <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">aqui</a> ou continuar usando o modelo padrão.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. O padrão regex "{value}" para {key} inválido, revertendo para o padrão. From d3588dd0069b71185e8a276c882baceb6102155f Mon Sep 17 00:00:00 2001 From: Voxel <41875513+VoxelCubes@users.noreply.github.com> Date: Wed, 7 Jan 2026 19:21:26 +0100 Subject: [PATCH 18/18] New translations panelcleaner_source.ts (Thai) --- translations/PanelCleaner_th_TH.ts | 995 +++++++++++++++-------------- 1 file changed, 500 insertions(+), 495 deletions(-) diff --git a/translations/PanelCleaner_th_TH.ts b/translations/PanelCleaner_th_TH.ts index 693a1f4..8769ab2 100644 --- a/translations/PanelCleaner_th_TH.ts +++ b/translations/PanelCleaner_th_TH.ts @@ -4,7 +4,7 @@ - + default reserved profile name ค่าเริ่มต้น @@ -20,54 +20,54 @@ สร้างโปรไฟล์ {profile_name} แล้ว - + Profile name cannot be empty. ชื่อของโปรไฟล์ไม่สามารถเว้นว่างได้ - + Profile name already in use. ชื่อของโปรไฟล์ได้ถูกใช้ไปแล้ว - + Profile name is reserved. ชื่อของโปรไฟล์ถูกสงวนไว้ - - - - + + + + Unknown ไม่ทราบ - + RGB Color mode RGB - + CMYK Color mode CMYK - + Grayscale Color mode ขาวดำ - + 1-bit Color mode 1บิต - + Palette Color mode ชุดสี @@ -78,44 +78,44 @@ - + Error downloading file from url: {url} Failed to save the file to {save_path} เกิดข้อผิดพลาดในการดาวน์โหลดไฟล์จาก url: {url} ไม่สามารถเซฟไฟล์ไปที่ {save_path} ได้ - + Error downloading file from url: {url} The file content is different from expected. เกิดข้อผิดพลาดในการดาวน์โหลดไฟล์จาก url: {url} เนื้อหาของไฟล์แตกต่างจากที่ควรจะเป็น + - - - + + + - - - + + N/A ไม่มีข้อมูล - + OCR Analytics OCR Analytics - + Number of boxes: {num_boxes} | Number of small boxes: {num_small_boxes} ({small_box_ratio}) จำนวนของกล่อง: {num_boxes} | จำนวนของกล่องเล็ก: {num_small_boxes} ({small_box_ratio}) - + Number of removed boxes: {num_removed} ({removed_box_ratio} total, {removed_among_small_ratio} of small boxes) จำนวนของกล่องที่ถูกลบ: {num_removed} (ทั้งหมด {removed_box_ratio} กล่อง , จาก {removed_among_small_ratio} กล่อง) @@ -147,65 +147,65 @@ Removed bubbles: หน้า {path}: {text} - + Small boxes กล่องเล็ก - + Removed boxes กล่องที่ถูกลบ - - + + Failed ล้มเหลว - + Perfect สมบูรณ์ - + Total ทั้งหมด - + Mask Fitment Analytics วิเคราะห์ความพอดีของ Mask - + Total ทั้งหมด - + Total boxes กล่องทั้งหมด - + Masks succeeded Masks สำเร็จ - + Masks failed ล้มเหลวในการ Masks - + Perfect masks Mask สมบูรณ์แบบ - + Average border deviation ค่าเบี่ยงเบนขอบเฉลี่ย @@ -222,7 +222,7 @@ Mask usage by thickness (in pixels): Mask ({thickness}px) - + Box mask Mask กล่อง @@ -240,27 +240,27 @@ Pages with failures / total: หน้าที่มีข้อผิดพลาด /ทั้งหมด: - + Denoising Analytics วิเคราะห์การปรับคุณภาพ - + Total masks Mask ทั้งหมด - + Masks denoised Mask ถูกปรับคุณภาพแล้ว - + Minimum deviation to denoise ค่าเบี่ยงเบนขั้นต่ำในการปรับคุณภาพ - + Maximum allowed deviation ค่าเบี่ยงเบนสูงสุดในการปรับคุณภาพ @@ -277,27 +277,27 @@ Pages with failures / total: ปรับคุณภาพ - + Inpainting Analytics วิเคราะห์การวาดทับ - + Inpainting performed การวาดทับเสร็จสมบูรณ์ - + Average thickness ความหนาโดยเฉลี่ย - + Minimum thickness ความหนาขั้นต่ำ - + Maximum thickness ความหนาสูงสุด @@ -354,7 +354,7 @@ Pages with failures / total: ไม่รู้จักไฟล์ชนิดนี้ - + The CSV file must start with a header row, followed by data rows. ไฟล์ CSV จะต้องเริ่มต้นด้วยส่วนหัวของแถวและตามด้วยแถวข้อมูล @@ -384,7 +384,7 @@ Pages with failures / total: การวิเคราะห์หลายรายการตรงกับเส้นทางของไฟล์ :{path} - + Error, no version found. ผิดพลาด,ไม่สามารถหาเวอร์ชั่นได้ @@ -394,7 +394,7 @@ Pages with failures / total: มีข้อผิดพลาดที่CUDA - + <html>Your GPU does not support the required CUDA operations.<br><br>Try uninstalling the current versions of torch and torchvision and installing the CPU version (or a different CUDA version) instead.<br>You can find further instructions here: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br>Check the "Compute Platform" section to see the available versions.<br><br>Your current CUDA version is: {cuda_version}<br></html> <html>GPU ของคุณไม่รอบรับ CUDA .<br><br>ลองลบ torch และ torchvision แล้วติดตั้งเวอร์ชั่น CPU หรือ CUDA เวอร์ชั่นอื่นแทน <br>คุณสามารถหาคำแนะนำได้ที่นี่: <br><a href='https://pytorch.org/get-started/locally/'>https://pytorch.org/get-started/locally/</a><br> ตรวจสอบ "ส่วนของ Compute Platform" เพื่อที่จะดูว่าต้องใช้เวอร์ชั่นไหน <br><br> เวอร์ชั่น CUDA ปัจจุบันของคุณคือ {cuda_version}<br></html> @@ -404,61 +404,61 @@ Pages with failures / total: หาโมเดลสำหรับการวาดทับไม่พบ - + filename ชื่อไฟล์ - + startx จุดเริ่มต้น x - + starty จุดเริ่มต้น y - + endx จุดจบ x - + endy จุดจบ y - + text ตัวหนังสือ - + Saved detected text to {output_file} เซฟข้อความที่พบไปที่ {output_file} - + Failed to write detected text to {output_file} ไม่สามารถเซฟข้อความที่พบไปที่ {output_file} ได้ - + Save Failed ไม่สามารถเซฟได้ - + Failed to write detected text to file. ไม่สามารถเซฟข้อความที่พบลงในไฟล์ได้ @@ -516,7 +516,7 @@ Pages with failures / total: - + Page layered export @@ -602,7 +602,7 @@ An AI-powered tool to clean manga panels. ColorButton - + Select Color เลือกสี @@ -648,7 +648,7 @@ An AI-powered tool to clean manga panels. ระบบไม่รองรับ - + This system is not supported. ระบบไม่รองรับ @@ -663,12 +663,12 @@ An AI-powered tool to clean manga panels. ถอนการติดตั้ง Dolphin - + The extension will be installed at {path}. This doesn't require root permissions. ส่วนขยายจะถูกติดตั้งที่ {path} การติดตั้งนี้ไม่ต้องการการอนุญาติจากผู้ดูแลระบบ - + The extension file will be deleted. This doesn't require root permissions. ส่วนขยายจะถูกลบออก การดำเนินการนี้ไม่ต้องการการอนุญาติจากผู้ดูแลระบบ @@ -683,17 +683,17 @@ An AI-powered tool to clean manga panels. ถอนการติดตั้งสำหรับ Windows Explorer - + The extension will be installed in the Windows registry.This requires admin permissions. ส่วนขยายจะถูกติดตั้งที่ Windows registry การดำเนินการนี้ต้องการการอนุญาติจากผู้ดูแลระบบ - + The extension will be uninstalled from the Windows registry.This requires admin permissions. ส่วนขยายจะถูกลบออกจาก Windows registry การดำเนินการนี้ต้องการการอนุญาติจากผู้ดูแลระบบ - + Dolphin extension installed successfully. ส่วนขยาย Dolphin ติดตั้งเสร็จสมบูรณ์ @@ -703,7 +703,7 @@ An AI-powered tool to clean manga panels. การติดตั้งเสร็จสมบูรณ์ - + Dolphin extension installed successfully. You may need to log out and log back in (or restart) for the changes to take effect. ส่วนขยาย Dolphin ติดตั้งเสร็จสมบูรณ์ @@ -783,12 +783,12 @@ By default, this uses the built-in profile. However, you can change this default FileTable - + Original size in pixels (width × height) ค่าดั้งเดิมในหน่วย pixel (กว้าง x ยาว) - + Processing size in pixels (width × height), scale factor ขนาดของการประมวลผลในหน่วย pixel (กว้าง x ยาว) @@ -814,7 +814,7 @@ By default, this uses the built-in profile. However, you can change this default ไม่รองรับรูปแบบ TIFF - + The following 5-channel TIFF files are not supported: ไม่รองรับรูปแบบ TIFF @@ -840,27 +840,27 @@ By default, this uses the built-in profile. However, you can change this default split - + Number of boxes removed by the OCR model / total boxes Number of boxes removed by the OCR model / total boxes - + Number of boxes that failed to generate a mask / total boxes Number of boxes that failed to generate a mask / total boxes - + Number of boxes that were perfectly masked / total boxes Number of boxes that were perfectly masked / total boxes - + Number of boxes that were denoised / total boxes Number of boxes that were denoised / total boxes - + Number of boxes that were inpainted Number of boxes that were inpainted @@ -1008,14 +1008,14 @@ By default, this uses the built-in profile. However, you can change this default Impossible Request - + Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. Denoising is disabled in the current profile, this output can't be generated. Please enable denoising in the profile settings and try again. - + Inpainting is disabled in the current profile, this output can't be generated. Please enable inpainting in the profile settings and try again. Inpainting is disabled in the current profile, this output can't be generated. @@ -1037,7 +1037,7 @@ Please enable inpainting in the profile settings and try again. Export failed - + Failed to export image: Failed to export image: @@ -1080,7 +1080,7 @@ Please enable inpainting in the profile settings and try again. result - + results results @@ -1095,17 +1095,17 @@ Please enable inpainting in the profile settings and try again. new image - + new images new images - + Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. Matched {num_images} {plural_images} to OCR results. {num_images_unassigned} {plural_img_unassigned}. - + {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. {num_analytics_unassigned} orphaned OCR {plural_result} will be lost. @@ -1204,12 +1204,12 @@ Please enable inpainting in the profile settings and try again. Note: Name "{name}" was hidden - + Corrupted log session Corrupted log session - + Current session Current session @@ -1224,7 +1224,7 @@ Please enable inpainting in the profile settings and try again. Error - + Errors Errors @@ -1234,7 +1234,7 @@ Please enable inpainting in the profile settings and try again. Critical - + Criticals Criticals @@ -1250,49 +1250,49 @@ Please enable inpainting in the profile settings and try again. MainWindow - + Open Config Statusbar button Open Config - + Open Log Statusbar button Open Log - + CUDA Enabled CUDA Enabled - + Select Output Directory Select Output Directory - + Aborting... Aborting... - + Loading OCR model... Loading OCR model... - + Loaded OCR model. Loaded OCR model. - + Error Error - + Another instance of Panel Cleaner appears to be running already.Opening a new instance will make the old session unstable. Continue anyway? @@ -1301,7 +1301,7 @@ Continue anyway? Continue anyway? - + Delete Models Delete Models @@ -1311,23 +1311,23 @@ Continue anyway? Are you sure you want to delete the machine learning models? This will make cleaning and OCR impossible until they are downloaded again. - + Encountered error: Encountered error: - + MPS Enabled MPS Enabled - - + + Inpainting Model Update Inpainting Model Update - + A new version of the inpainting model is available. You can delete the model later if you don't want to upgrade yet. Switch to the new model? @@ -1341,148 +1341,148 @@ Switch to the new model? Old model kept. To upgrade, select "Help" then "Delete Machine Learning Models" from the menubar. - + Layout Reset Layout Reset - + The window layout has been reset. The default layout will be used next time. The window layout has been reset. The default layout will be used next time. - + Multiple Instances Multiple Instances - + Found process ID in lock file: Found process ID in lock file: - + Failed to delete models. Failed to delete models. - + Delete Failed Delete Failed - + Models Deleted Models Deleted - + The models were deleted. Would you like to download them again? The models were deleted. Would you like to download them again? - + Not Supported Not Supported - + Unsupported system or package format Unsupported system or package format - + Shutdown Shutdown - + After processing, the system will shut down. After processing, the system will shut down. - + You have review options enabled, these will need to be manually closed before the action can start. You have review options enabled, these will need to be manually closed before the action can start. - + You have disabled writing output. You have disabled writing output. - + Shutting down system... Shutting down system... - + Running post action "{action}"... Running post action "{action}"... - + Restart Required Restart Required - + The language has been changed. Please restart the application for the changes to take effect. The language has been changed. Please restart the application for the changes to take effect. - - + + System Language System Language - + Import Profile Import Profile - + Profile Files (*.conf) Profile Files (*.conf) - + Profile Imported Profile Imported - + Import Error Import Error - + Failed to Delete Failed to Delete - + The default profile cannot be deleted. The default profile cannot be deleted. - + Delete Profile Delete Profile - + Are you sure you want to delete the profile {profile_name}? Are you sure you want to delete the profile {profile_name}? - + Delete Error Delete Error - + Failed to delete the profile. Failed to delete the profile. @@ -1494,217 +1494,217 @@ Switching profiles will discard changes to the current profile. Switching profiles will discard changes to the current profile. - + Unsaved changes Unsaved changes - - + + Load Error Load Error - - + + Failed to load profile: Failed to load profile: - + Tesseract OCR is not installed or not found Tesseract OCR is not installed or not found - + <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> <html>Can't use Tesseract to perform OCR. Reverting to manga-ocr. Please see the instructions to install Tesseract correctly <a href="https://github.com/VoxelCubes/PanelCleaner?tab=readme-ov-file#ocr">here</a> or continue using the default model.</html> - + Unsupported Language Unsupported Language - + The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. The language '{lang_name}' is not supported by any of your current OCR engines. Check the online documentation for how to add support for more languages. - + Profile Created Profile Created - + Create Error Create Error - + Failed to save profile to {profile_path} Failed to save profile to {profile_path} - - + + Save Error Save Error - + Failed to save profile. Failed to save profile. - + Profile saved to {profile_path} Profile saved to {profile_path} - + Failed to save config. Failed to save config. - + Failed to save the new profile to the configuration file. Failed to save the new profile to the configuration file. - - + + No Files No Files - + No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. No files to process. You can add files by dragging and dropping them in the middle of the window, or through the menubar: Files -> Add Files or Add Folder. - + No Outputs No Outputs - + No outputs were requested. Please select at least one output before cleaning. No outputs were requested. Please select at least one output before cleaning. - - + + cleaned cleaned - + No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. No files to process. To edit an old OCR output file, you must first load (one or more of) the images to which it corresponds. - + Open OCR Output File Open OCR Output File - + OCR Output Files (*.txt *.csv) OCR Output Files (*.txt *.csv) - + Parse Error Parse Error - + Failed to parse the OCR output file. Failed to parse the OCR output file. - + File Exists File Exists - + The file '{output_path}' already exists. Overwrite? The file '{output_path}' already exists. Overwrite? - + Export Images Export Images - + Would you like to export the cleaned images? Would you like to export the cleaned images? - + Export OCR Results Export OCR Results - + After processing, the "{action}" action will be executed. After processing, the "{action}" action will be executed. - - + + Post action canceled. Post action canceled. - + Un-Applied Changes Un-Applied Changes - + You have un-applied changes to the profile. Continue anyway? You have un-applied changes to the profile. Continue anyway? - + Would you like to export the OCR results? Would you like to export the OCR results? - + Processing Finished Processing Finished - + Finished processing all files. Finished processing all files. - + Processing Aborted Processing Aborted - + Processing aborted. Processing aborted. - + Processing Error Processing Error - + Encountered an error while processing files. Encountered an error while processing files. @@ -1715,7 +1715,7 @@ Check the online documentation for how to add support for more languages. - + Profile Profile @@ -1730,360 +1730,365 @@ Check the online documentation for how to add support for more languages.Apply - - + + Save Save - + Reset All Reset All - + Drag and Drop Images or Folders Here Drag and Drop Images or Folders Here - + Images Images - - + + File File - + Size Size - + Processing Size The size that the picture is shrunk down to for processing. Processing Size - + File Size File Size - + Color Mode Like RGB, grayscale etc. Color Mode - + Analytics Can also call this statistics. Analytics - + Process Process - + Cleaning Cleaning - + OCR OCR - + Edit Existing Output Edit Existing Output - + Abort Abort - + Start Start - + Output Output - + Cleaned Image Cleaned Image - + Mask Mask - + Isolated Text Isolated Text - - + + Review Output Review Output - + <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> <html><head/><body><p>When checked, a review window will open upon process completion. The results are also pre-viewable in each image's details view, before proceeding with saving them to disk.</p></body></html> - + Save Output Save Output - + Output Directory: Output Directory: - + You can use a relative path to create a subfolder at the image's original location, or use an absolute path. The &apos; is the possessive 's apostrophe and should be ignored if the language doesn't call for it. You can use a relative path to create a subfolder at the image's original location, or use an absolute path. - + Plain Text Plain Text - + When checked, a review window will open upon process completion. When checked, a review window will open upon process completion. - + Output File: Output File: - + <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> <html><head/><body><p>Either enter an absolute or relative path with a file name. The output of all images is written to the same file.</p></body></html> - + detected_text.txt The file suffix .txt is important, don't change that. The .csv extension is swapped in automatically for that alternative when selected. detected_text.txt - + Cancel Action Cancel Action - + Close Close - + Generating: This is present progressive, as in "[currently] Generating:" Generating: - + Current Step: Current Step: - - + + %v / %m %v / %m - + Images Processed: As in "images processed: 42" with progress bar. Images Processed: - + Set Default Set Default - + Settings Settings - + Theme As in color theme Theme - + Language Language - + Help Help - + Add Files... Add Files... - + Remove All Files Remove All Files - + Remove File Remove File - + Help Translate Panel Cleaner Help Translate Panel Cleaner - + Report an Issue... Report an Issue... - + Simulate Exception Simulate Exception - + Add File Browser Extension... Add File Browser Extension... - + Show Supported OCR Languages Show Supported OCR Languages - - + + Reset Window Layout Preferences Reset Window Layout Preferences - + Show Out Of Memory Warnings Show Out Of Memory Warnings - + + Clear Image Cache When Closing + Clear Image Cache When Closing + + + Post-Run Actions Post-Run Actions - + Configure Post-Run Actions... Configure Post-Run Actions... - + New New - + <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> <html><head/><body><p>When checked, the outputs are saved on disk. Otherwise, you can only preview them in the image details view or the Output Review, if enabled. </p><p>If the profile remains unchanged after cleaning without this option enabled, you can quickly export them by running the cleaner again with this option enabled.</p></body></html> - + CSV Spreadsheet CSV Spreadsheet - + Delete Delete - + Import... Import... - + Online Documentation Online Documentation - + About Panel Cleaner About Panel Cleaner - + View License View License - + Save as... Save as... - + System System - + Dark Dark - + Light Light - + Add Folder... Add Folder... - + Delete Machine Learning Models Delete Machine Learning Models - + Download Machine Learning Models Download Machine Learning Models - + Donate Donate @@ -2115,19 +2120,19 @@ Check the online documentation for how to add support for more languages.OCR model: - + Already downloaded Already downloaded - + Text Detector model (CUDA) Text Detector model (CUDA) - + Text Detector model (CPU) Text Detector model (CPU) @@ -2146,7 +2151,7 @@ Check the online documentation for how to add support for more languages.OCR download errors - + ETA estimated time of completion ETA @@ -2724,7 +2729,7 @@ Check the online documentation for how to add support for more languages.The name "{shutdown_command}" is reserved. - + The Flatpak sandbox will likely prevent actions from working. The Flatpak sandbox will likely prevent actions from working. @@ -2805,73 +2810,73 @@ Check the online documentation for how to add support for more languages. Process Steps - + The original image with scaling applied (if needed). Step description in the image details view, step: Input, output: Input The original image with scaling applied (if needed). - + Input Step name in the image details view Input - + Input Output name in the image details view Input - + The rough mask generated by the AI. Step description in the image details view, step: Text Detection, output: AI Mask The rough mask generated by the AI. - + Text Detection Step name in the image details view Text Detection - + AI Mask Output name in the image details view AI Mask - + The unfiltered box data generated by the AI. Step description in the image details view, step: Text Detection, output: Raw Boxes The unfiltered box data generated by the AI. - + Raw Boxes Output name in the image details view Raw Boxes - + The outlines of the text boxes the AI found. Step description in the image details view, step: Preprocessor, output: Initial Boxes The outlines of the text boxes the AI found. - + Preprocessor Step name in the image details view Preprocessor - + Initial Boxes Output name in the image details view Initial Boxes - + The final boxes after expanding, merging and filtering unneeded boxes with OCR. Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. Step description in the image details view, step: Preprocessor, output: Final Boxes @@ -2879,181 +2884,181 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: reference boxes for denoising. - + Final Boxes Output name in the image details view Final Boxes - + The mask of the merged boxes. Step description in the image details view, step: Masker, output: Box Mask The mask of the merged boxes. - + Masker Step name in the image details view Masker - + Box Mask Output name in the image details view Box Mask - + The rough text detection mask with everything outside the box mask cut out. Step description in the image details view, step: Masker, output: Cut Mask The rough text detection mask with everything outside the box mask cut out. - + Cut Mask Output name in the image details view Cut Mask - + The different steps of growth around the cut mask displayed in different colors. Step description in the image details view, step: Masker, output: Mask Layers The different steps of growth around the cut mask displayed in different colors. - + Mask Layers Output name in the image details view Mask Layers - + The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). Step description in the image details view, step: Masker, output: Fitment Quality The standard deviation (σ) and outline thickness (in pixels) of each best mask chosen, if any. Lower σ is better, from perfect (purple) to failed (red). - + Fitment Quality Output name in the image details view Fitment Quality - + The input image with the final mask overlaid in color. Step description in the image details view, step: Masker, output: Mask Overlay The input image with the final mask overlaid in color. - + Mask Overlay Output name in the image details view Mask Overlay - + The collection of masks for each bubble that fit best. Step description in the image details view, step: Masker, output: Final Mask The collection of masks for each bubble that fit best. - + Final Mask Output name in the image details view Final Mask - + The text layer isolated from the input image. Step description in the image details view, step: Masker, output: Isolated Text The text layer isolated from the input image. - + Isolated Text Output name in the image details view Isolated Text - + The input image with the final mask applied. Step description in the image details view, step: Masker, output: Masked Output The input image with the final mask applied. - + Masked Output Output name in the image details view Masked Output - + The masks that required denoising, to be overlaid on the final mask when exporting. Step description in the image details view, step: Denoiser, output: Denoise Mask The masks that required denoising, to be overlaid on the final mask when exporting. - + Denoiser Step name in the image details view Denoiser - + Denoise Mask Output name in the image details view Denoise Mask - + The input image with the denoised mask applied. Step description in the image details view, step: Denoiser, output: Denoised Output The input image with the denoised mask applied. - + Denoised Output Output name in the image details view Denoised Output - + The inpainted sections around the text that was poorly cleaned, if at all. Step description in the image details view, step: Inpainter, output: Inpainted Mask The inpainted sections around the text that was poorly cleaned, if at all. - + Inpainter Step name in the image details view Inpainter - + Inpainted Mask Output name in the image details view Inpainted Mask - + The input image: cleaned, denoised (if enabled), and inpainted. Step description in the image details view, step: Inpainter, output: Inpainted Output The input image: cleaned, denoised (if enabled), and inpainted. - + Inpainted Output Output name in the image details view Inpainted Output - + <This is just a placeholder, translating this string is not necessary> Step description in the image details view, step: None, output: Write Output <This is just a placeholder, translating this string is not necessary> - + Write Output Output name in the image details view Write Output @@ -3062,1141 +3067,1141 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r Profile - + General Profile section title General - + About this profile: Profile comment in General About this profile: - + Notes Profile entry in General Notes - + Preferred file type to save the cleaned image as. Profile comment in General Preferred file type to save the cleaned image as. - + Preferred File Type Profile entry in General Preferred File Type - + Preferred file type to save the mask as. Only image formats that allow for transparency are supported. Profile comment in General Preferred file type to save the mask as. Only image formats that allow for transparency are supported. - + Preferred Mask File Type Profile entry in General Preferred Mask File Type - + The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. Profile comment in General The following are the lower and upper targets for the height of the input image. It is only ever scaled down to fit within the range, preferring whole number factors to minimize the impact on image quality. Images smaller than either target will remain unchanged. You can disable this feature by setting one or both values less than or equal to 0. - + This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. Profile comment in General This is useful for significantly speeding up processing on large images. Also, since other options relying on pixel dimensions depend on size, this will help normalize the results across different image sizes. - + The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. Profile comment in General The image will be scaled down, processed, and then only the mask is scaled back up. Meaning that the cleaned output will still use the original, unscaled image to prevent any loss in quality. Only the height of the image is used to determine the scale factor, preserving the aspect ratio, and ignoring the individual width of an image so that the factor remains consistent if one of the pages is a double page spread. - + E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. Profile comment in General E.g. for a lower target of 1000 and an upper target of 2000, an image with the size 5000x7000 (w, h) pixels will be scaled down by a factor of 4, so that it has the size 1250x1750 pixels during processing. - + Input Height Lower Target Profile entry in General Input Height Lower Target - + Input Height Upper Target Profile entry in General Input Height Upper Target - + Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). Profile comment in General Split long strips into individual pages. If enabled, instead of squeezing the entire strip to fit into the preferred height, the strip will be split into individual pages, each fitting the preferred height plus/minus double the split tolerance margin (if the segment was in the middle of the strip). - + Split Long Strips Profile entry in General Split Long Strips - + Preferred height to split long strips at. Profile comment in General Preferred height to split long strips at. - + Preferred Split Height Profile entry in General Preferred Split Height - + Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. Profile comment in General Tolerance margin for splitting long strips. This is the maximum difference between the preferred split height and the actual height of the strip. An algorithm determines the best split point within this margin, in an effort to avoid splitting in the middle of a panel. - + Split Tolerance Margin Profile entry in General Split Tolerance Margin - + Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. Profile comment in General Aspect ratio to use for splitting long strips. This is the ratio of the width to the height of the image. If the image's aspect ratio is smaller than this value, it will be considered a long strip.<br> Example: 0.3 means that the width of the image is 0.3 times the height. - + Long Strip Aspect Ratio Profile entry in General Long Strip Aspect Ratio - + Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. Profile comment in General Merge long strips back into a single image. If enabled, the individual pages created from a long strip will be merged back into a single image upon export. - + Merge After Split Profile entry in General Merge After Split - + Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in General Maximum number of threads to use for exporting images. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Export Profile entry in General Max Threads Export - + Text Detector Profile section title Text Detector - + Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Path to the text detection model, leave empty to use the built-in model. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> Profile comment in Text Detector Path to the text detection model, leave empty to use the built-in model. It is only recommended to override this if the version downloaded automatically is older than the latest release. The path must point directly to the comictextdetector.pt (CUDA) or comictextdetector.pt.onnx (CPU) file. You can download older versions of the model <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here.</a> - + Model Path Profile entry in Text Detector Model Path - + Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. DO NOT report issues about this problem, as it's very hardware-dependent! Profile comment in Text Detector Number of models to run in parallel. This is useful if you have enough RAM (or VRAM with CUDA) to run multiple models at the same time. This, of course, will increase the speed of the process, but can also crash your computer if you overestimate your hardware. I recommend using 1 model per 2 GB of memory available, or 1 model per 50 images. Note: This is ignored if processing less than 50 files due to the overhead of starting multiple models not being worth it below that. Warning: This may cause program instability, use at your own risk. <br>DO NOT report issues about this setting, as it's entirely hardware-dependent! - + Concurrent Models Profile entry in Text Detector Concurrent Models - + Preprocessor Profile section title Preprocessor - + Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. Profile comment in Preprocessor Box sizes are given in the total number of pixels, so a box of 200x200 pixels has a size of 200 * 200 = 40000. - + Minimum size of any box to keep it. Profile comment in Preprocessor Minimum size of any box to keep it. - + Box Min Size Profile entry in Preprocessor Box Min Size - + Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. Profile comment in Preprocessor Minimum size of a box with "unknown" language to keep it. This language is typically assigned to logos and other badly behaved text elements. - + Suspicious Box Min Size Profile entry in Preprocessor Suspicious Box Min Size - + The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. Profile comment in Preprocessor The minimum overlap between two boxes to merge them. This percentage (0-100) means how much of the smaller box must be inside the larger box to be merged. A higher value will require a larger overlap to merge the boxes. - + Box Overlap Threshold Profile entry in Preprocessor Box Overlap Threshold - + Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. Profile comment in Preprocessor Whether to use OCR to detect boxes that aren't worth cleaning, like ones that only contain numbers or symbols. - + OCR Enabled Profile entry in Preprocessor OCR Enabled - + Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. Profile comment in Preprocessor Whether to use Tesseract to perform OCR tasks.<br> If checked, Tesseract OCR can be used for text extraction, if available.<br> If unchecked, the built-in OCR model (manga-ocr) is always used, which is best suited for vertical Japanese text. - + OCR Use Tesseract Profile entry in Preprocessor OCR Use Tesseract - + The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. Profile comment in Preprocessor The language to use for OCR tasks. The text detector can detect Japanese and English, for anything else, select a language explicitly here. Detecting per box retains what the text detector detected, while per page will assign the most prominent language to the entire page.<br> Note: Only Japanese is supported out of the box, everything else requires Tesseract to be enabled and its associated language packs to be installed. - + OCR Language Profile entry in Preprocessor OCR Language - + Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. Profile comment in Preprocessor Specifies which engine to use for performing OCR.<br> - auto: Automatically selects the OCR engine based on the detected language of each text block within the image. Uses Manga Ocr for Japanese text, Tesseract for English or Unknown Text.<br> - mangaocr: Forces Panel Cleaner to use the built-in manga-ocr model for all text recognition tasks. Best suited for vertical Japanese text.<br> - tesseract: Forces Panel Cleaner to use Tesseract OCR for all text recognition tasks. This is a versatile option that supports English and multiple other languages. - + OCR Engine Profile entry in Preprocessor OCR Engine - + Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. Profile comment in Preprocessor Defines the reading order for processing and sorting text boxes on the entire page, not individual text blocks. This global setting influences how text boxes are ordered and presented for further processing.<br> - auto: Detects the reading order based on the detected language of each text block within the page.<br> - manga: Right-to-left, top-to-bottom order. Suitable for Japanese manga.<br> - comic: Left-to-right, top-to-bottom order. Suitable for Western comics and texts.<br> Choose based on the predominant layout of your content. - + Reading Order Profile entry in Preprocessor Reading Order - + Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. Profile comment in Preprocessor Maximum size of a box to perform OCR on. These useless boxes are usually small, and OCR is slow, so use this as a cutoff. - + OCR Max Size Profile entry in Preprocessor OCR Max Size - + Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the OCR model returns full-width characters, so this pattern should match them. Profile comment in Preprocessor Regex pattern to match against OCR results. Anything matching this pattern is discarded.<br> Note: the MangaOCR model returns full-width characters, so this pattern should match them. - + OCR Blacklist Pattern Profile entry in Preprocessor OCR Blacklist Pattern - + The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. The standard OCR model can only handle Japanese text, so when strict is enabled, it will discard boxes that the it isn't confident are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. Profile comment in Preprocessor The MangaOCR model can only handle Japanese text, so when strict is enabled, it will discard boxes where the Text Detector isn't confident that they are Japanese. Sometimes, numbers or other symbols will lower its confidence, resulting in the detected language being unknown. If strict is disabled, those will not be discarded. Anything that is confidently recognized as a different language will be discarded regardless of this setting.<br> Note: this setting is only relevant when ocr_language is set to detect per box or page. - + OCR Strict Language Profile entry in Preprocessor OCR Strict Language - + Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the initial boxes created by the text detector AI. These boxes are visualized in green. - + Box Padding Initial Profile entry in Preprocessor Box Padding Initial - - + + Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. Profile comment in Preprocessor Padding to add to the right side of a box. This extension helps to cover rubytext that floats off to the right of vertical text. - + Box Right Padding Initial Profile entry in Preprocessor Box Right Padding Initial - + Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. Profile comment in Preprocessor Padding to add to each side of a box. This is added to an extended set of boxes, used to cut out false positives by the text detector AI's mask. These boxes are visualized in purple. - + Box Padding Extended Profile entry in Preprocessor Box Padding Extended - + Box Right Padding Extended Profile entry in Preprocessor Box Right Padding Extended - + Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. Profile comment in Preprocessor Padding to add to each side of a box. This is added to the reference boxes used to sample the original image while analyzing what mask to use. These boxes are visualized in blue. - + Box Reference Padding Profile entry in Preprocessor Box Reference Padding - + Masker Profile section title Masker - + Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Masker Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Masker Max Threads - + Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. Profile comment in Masker Number of pixels to grow the mask by each step. This bulks up the outline of the mask, so smaller values will be more accurate but slower. - + Mask Growth Step Pixels Profile entry in Masker Mask Growth Step Pixels - + Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. Profile comment in Masker Number of steps to grow the mask by. A higher number will make more and larger masks, ultimately limited by the reference box size. - + Mask Growth Steps Profile entry in Masker Mask Growth Steps - + Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. Profile comment in Masker Minimum thickness of a mask. This is like the first mask's growth step, the remaining will follow Mask Growth Step Pixels. This way, you can have a small step size for accuracy, but still prevent very thin masks that might be used to clean text only surrounded by an outline, when inpainting would've been the better choice. - + Min Mask Thickness Profile entry in Masker Min Mask Thickness - + The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. Profile comment in Inpainter The minimum standard deviation of colors around the edge of a given mask to perform inpainting on the region around the mask. If this value matches the maximum deviation for masks, only failed bubbles will be inpainted, making the following two settings irrelevant. - + Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. Profile comment in General Combine outputs into a single project file as layers. Currently supported formats: Photoshop PSD.<br> - none: Each image and mask are saved as basic files.<br> - psd-per-image: Images and masks are saved together in a PSD file per input image.<br> - psd-bulk: All images and masks are saved together in a single PSD file, grouped by input image. - + Layered Export Profile entry in General Layered Export - + Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. Profile comment in Masker Whether to allow colored masks. When true, the masker will allow masks to use any color, not just white, black, or gray. - + Allow Colored Masks Profile entry in Masker Allow Colored Masks - + Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). Profile comment in Masker Maximum threshold for a pixel to be considered off-white. The median color along the edge of a mask may not be pure white, so to prevent slight off-white tones on a pure-white background, anything lighter than this threshold value will be rounded up to pure white. Value range: black (0) to pure white (255). - + Off White Max Threshold Profile entry in Masker Off White Max Threshold - + The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Minimum improvement in standard deviation of the mask to continue shrinking it. The standard deviation refers to the variation is color along the edge of a mask. A low variation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. Setting a higher value here requires a higher improvement to consider a smaller mask, to give a preference to larger masks. Profile comment in Masker The standard deviation of a mask represents the variation in color along the edge of the mask. For this, only the single line of pixels along the outer edge of a mask is sampled from the original image. A low deviation means that the mask sits in a solid color, which means it doesn't intersect any text or other objects. - + Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Minimum improvement in standard deviation of the mask to continue shrinking it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. Profile comment in Masker Minimum improvement in standard deviation of the mask to continue growing it. Setting a higher value here requires a higher improvement to consider a larger mask, to give a preference to smaller masks. - + Mask Improvement Threshold Profile entry in Masker Mask Improvement Threshold - + Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. Profile comment in Masker Whether to use the fast mask selection algorithm. When true, the mask selection algorithm will pick the first perfect mask, if one is found early. This is faster, but may not find the best mask, if a slightly bigger one would have been better. - + Mask Selection Fast Profile entry in Masker Mask Selection Fast - + Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Maximum number of threads to use for mask generation. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. Profile comment in Denoiser Maximum number of threads to use for denoising. You can leave it unspecified to use all available threads. Lower this value if you run into memory issues, which will appear as random crashes. - + Max Threads Profile entry in Denoiser Max Threads - + Inpainter Profile section title Inpainter - + EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note tha inpainting isn't enabled by default. Profile comment in Inpainter EXPERIMENTAL FEATURE: If you find better default settings, please open an issue on github to share the improvements with everyone. Note that inpainting isn't enabled by default. - + Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. Profile comment in Inpainter Inpainting is when machine learning is used to replace the content of an image based on its surroundings. For masks that couldn't be cleaned well (or at all), inpainting can be used. To differentiate this from denoising, inpainting is meant for significantly worse masks that had a tight fit. Any masks that were denoised won't be inpainted. - + Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. Since this step can provide poor results in some cases, it can be disabled here. Set to False to disable inpainting. Profile comment in Inpainter Since this step can provide poor results in some cases, it can be disabled here. Uncheck to disable inpainting. - + Inpainting Enabled Profile entry in Inpainter Inpainting Enabled - + Inpainting Min Std Dev Profile entry in Inpainter Inpainting Min Std Dev - + The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. Profile comment in Inpainter The maximum radius of a mask to perform inpainting on. Masks larger than this will be left as they are, because if the margin is that big, it is likely that the mask is already good enough. - + Inpainting Max Mask Radius Profile entry in Inpainter Inpainting Max Mask Radius - + The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. Profile comment in Inpainter The minimum radius around a mask to inpaint. This is added to the optimal mask size to ensure that the inpainting covers the entire mask. - + Min Inpainting Radius Profile entry in Inpainter Min Inpainting Radius - + For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation Profile comment in Inpainter For masks that proved far harder to clean, meaning they had a high standard deviation, increase the radius of the inpainting to cover more of the mask. This is additional margin is added to the min inpainting radius and is calculated as: inpainting radius multiplier times mask standard deviation - + Inpainting Radius Multiplier Profile entry in Inpainter Inpainting Radius Multiplier - + The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. Profile comment in Inpainter The maximum radius around a mask to inpaint. This limits the size the inpainting can grow to, to prevent it from covering too much of the image, if a large radius multiplier is used. - + Max Inpainting Radius Profile entry in Inpainter Max Inpainting Radius - + After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. After inpainting, cut the result out of the original image to prevent the inpaitning from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. Profile comment in Inpainter After inpainting, cut the result out of the original image to prevent the inpainting from affecting the rest of the image. This ensures that the original image is preserved as much as possible. This radius is added around the final inpainting radius, due to the inpainting model modifying a few pixels outside of its dedicated region. - + Inpainting Isolation Radius Profile entry in Inpainter Inpainting Isolation Radius - + Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. Profile comment in Inpainter Fade the edges of the inpainted cover mask by this many pixels to smoothly blend the inpainted parts into the rest of the image.<br> If you see faint outlines after inpainting, increase the min inpainting value and set this one to half that amount. - + Inpainting Fade Radius Profile entry in Inpainter Inpainting Fade Radius - + Detect per box Language option for OCR Detect per box - + Detect per page Language option for OCR Detect per page - + Japanese Language option for OCR Japanese - + English Language option for OCR English - + Korean Language option for OCR Korean - + Korean (vertical) Language option for OCR Korean (vertical) - + Chinese - Simplified Language option for OCR Chinese - Simplified - + Chinese - Traditional Language option for OCR Chinese - Traditional - + Albanian Language option for OCR Albanian - + Arabic Language option for OCR Arabic - + Azerbaijani Language option for OCR Azerbaijani - + Azerbaijani - Cyrilic Language option for OCR Azerbaijani - Cyrilic - + Bengali Language option for OCR Bengali - + Bulgarian Language option for OCR Bulgarian - + Burmese Language option for OCR Burmese - + Catalan; Valencian Language option for OCR Catalan; Valencian - + Croatian Language option for OCR Croatian - + Czech Language option for OCR Czech - + Danish Language option for OCR Danish - + Dutch; Flemish Language option for OCR Dutch; Flemish - + Esperanto Language option for OCR Esperanto - + Estonian Language option for OCR Estonian - + Finnish Filipino Language option for OCR Finnish - + French Language option for OCR French - + Georgian Language option for OCR Georgian - + German Language option for OCR German - + Greek Language option for OCR Greek - + Hebrew Language option for OCR Hebrew - + Hindi Language option for OCR Hindi - + Hungarian Language option for OCR Hungarian - + Indonesian Language option for OCR Indonesian - + Italian Language option for OCR Italian - + Kazakh Language option for OCR Kazakh - + Latin Language option for OCR Latin - + Lithuanian Language option for OCR Lithuanian - + Luxembourgish Language option for OCR Luxembourgish - + Malay Language option for OCR Malay - + Mongolian Language option for OCR Mongolian - + Nepali Language option for OCR Nepali - + Norwegian Language option for OCR Norwegian - + Persian Language option for OCR Persian - + Polish Language option for OCR Polish - + Portuguese Language option for OCR Portuguese - + Romanian; Moldavian Language option for OCR Romanian; Moldavian - + Russian Language option for OCR Russian - + Serbian Language option for OCR Serbian - + Serbian - Latin Language option for OCR Serbian - Latin - + Slovak Language option for OCR Slovak - + Slovenian Language option for OCR Slovenian - + Spanish; Castilian Language option for OCR Spanish; Castilian - + Swedish Language option for OCR Swedish - + Tagalog Language option for OCR Tagalog - + Tamil Language option for OCR Tamil - + Telugu Language option for OCR Telugu - + Thai Language option for OCR Thai - + Turkish Language option for OCR Turkish - + Ukrainian Language option for OCR Ukrainian - + Vietnamese Language option for OCR Vietnamese - + The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. Profile comment in Masker The maximum standard deviation of a mask to consider. A high value here means a higher tolerance for the mask intersecting text or other objects, which isn't a good mask, as it will require inpainting anyway. Setting this to 0 will only allow perfect masks, which is recommended for very high resolution images. - + Mask Max Standard Deviation Profile entry in Masker Mask Max Standard Deviation - + Color to use for the debug mask. Profile comment in Masker Color to use for the debug mask. - + Debug Mask Color Profile entry in Masker Debug Mask Color - + Denoiser Profile section title Denoiser - + When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. Profile comment in Denoiser When a bit of text is hard to mask off, the cleaning step likely had to choose a small mask, which leaves a lot of jpeg-artifacts behind, if they were around the text. - + This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. Profile comment in Denoiser This means that the color of pixels around the edge of a mask isn't uniform, which is quantified as a standard deviation. The denoiser can target masks with a minimum standard deviation and denoise the area right around the mask. This preserves details in the rest of the image, but removes artifacts right around where the text used to be. - + Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. Profile comment in Denoiser Since this is an optional step and may even be superfluous for high-resolution images that don't suffer from jpeg-artifacts, it can be disabled here. Set to False to disable denoising. - + Denoising Enabled Profile entry in Denoiser Denoising Enabled - + The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. Profile comment in Denoiser The minimum standard deviation of colors around the edge of a given mask to perform denoising on the region around the mask. - + Noise Min Standard Deviation Profile entry in Denoiser Noise Min Standard Deviation - + Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. Profile comment in Denoiser Note: If inpainting is enabled, the inpainting min std deviation will act as a maximum for this, after which this mask is handed off to the inpainter. - + The thickness of an outline to denoise around a mask. Profile comment in Denoiser The thickness of an outline to denoise around a mask. - + Noise Outline Size Profile entry in Denoiser Noise Outline Size - + Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. Profile comment in Denoiser Fade the edges of the denoised cover mask by this amount to smoothly blend the denoised parts into the rest of the image. - + Noise Fade Radius Profile entry in Denoiser Noise Fade Radius - + Set to True to support colored images. Profile comment in Denoiser Set to True to support colored images. - + Colored Images Profile entry in Denoiser Colored Images - + How strongly to denoise the image. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the image. Higher values mean more denoising. - + Filter Strength Profile entry in Denoiser Filter Strength - + How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. Profile comment in Denoiser How strongly to denoise the color channels, as opposed to lightness. Higher values mean more denoising. - + Color Filter Strength Profile entry in Denoiser Color Filter Strength - + Size in pixels of the template patch that is used to compute weights. Should be odd. Profile comment in Denoiser Size in pixels of the template patch that is used to compute weights. Should be odd. - + Template Window Size Profile entry in Denoiser Template Window Size - + Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. Profile comment in Denoiser Size in pixels of the window that is used to compute weighted average for given pixel. Should be odd. - + Search Window Size Profile entry in Denoiser Search Window Size @@ -4223,7 +4228,7 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r px² - + All Cores The limit on the number of CPU cores to use, or here, no limit. All Cores @@ -4640,14 +4645,14 @@ Green: initial boxes. Red: extended boxes. Purple: merged (final) boxes. Blue: r ProfileToolBox - + <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> <html>The Text Detector model path {value} does not exist, reverting to default. You can download the model manually from <a href="https://github.com/zyddnys/manga-image-translator/releases/latest">here</a> or continue using the default model.</html> - + The regex pattern "{value}" for {key} is invalid, reverting to default. The regex pattern "{value}" for {key} is invalid, reverting to default.