🌸 Flower Classification App (Flutter)

• Flower Classification App is a cross-platform Flutter application that identifies flower species from images using an offline Machine Learning model.
• Users can capture a photo using the camera or select one from the gallery, and the app predicts the flower entirely on-device using TensorFlow Lite.
• The app works 100% offline, ensuring fast inference, user privacy, and no internet dependency.

---

🚀 Key Features

Image Input

• Capture flower images using the device camera
• Select images from the photo gallery

Offline AI Prediction

• On-device image classification
• No cloud APIs, no network calls
• Displays:
  • Predicted flower name
  • Confidence percentage

Flower Information Panel

• Displays basic information about the predicted flower
• Smooth animated info cards
• Clean separation between prediction result and flower details

Modern UI & UX

• Vibrant Material-based design
• Smooth animations for result & info card
• Supports multy color themes
• Optimized for mobile performance

Supported Flower Categories

• Trained on 50 flower classes
• Based on a Kaggle 299 Flowers Dataset
• Scalable architecture (easy to add more classes)

---

🤖 Machine Learning: Offline Image Classification

The app uses a Convolutional Neural Network (CNN) with Transfer Learning to achieve high accuracy while remaining lightweight for mobile devices.

Model Details

• Architecture: MobileNetV2 (Transfer Learning)
• Framework: TensorFlow / Keras
• Inference Engine: TensorFlow Lite
• Execution: Fully offline (on-device)

Dataset Used

• Kaggle 299 Flowers Dataset
• 50 flower classes
• Custom split into:
  • train
  • val
  • test
• Preprocessed and trained using Google Colab

ML Workflow

• Dataset prepared and split offline
• Model trained in Google Colab
• Converted to TensorFlow Lite (.tflite)
• Integrated into Flutter using tflite_flutter
• Real-time inference on mobile device

---

🛠️ Technologies Used

App Development

• Flutter (Dart)

Machine Learning

• TensorFlow
• TensorFlow Lite
• CNN with Transfer Learning (MobileNetV2)

Flutter Packages

• tflite_flutter
• image_picker
• image

Tools

• Android Studio
• VS Code
• Google Colab
• Flutter SDK

---

📦 Installation & Setup

1. Clone this repository:

   git clone https://github.com/Kalana-S/Flower-Classification-App.git
   cd Flower-Classification-App
   

2. Get all dependencies:

   flutter pub get
   

3. Run the app:

   flutter run
   

4. (Optional) Build a release APK:

   flutter build apk --release
   

---

💡 How to Use

1. Launch the app
2. Select Camera or Gallery.
3. Capture or select a flower image.
4. The app predicts:
   • Predicted flower
   • Confidence score
   • Flower information
5. Results appear instantly — no internet required.

---

🎥 App Demo (Screen Recording)

Full app workflow — UI → Input → Offline Prediction

video_2.webm

---

🔒 Offline & Privacy-Friendly

• Works completely offline
• No data collection
• No images uploaded to servers
• All ML inference runs locally on the device

---

🤝 Contribution

Contributions are welcome.
Feel free to fork the repository and submit a pull request for:

• UI enhancements
• Performance optimizations
• Additional flower data
• New features

---

📜 License

This project is open-source and available under the MIT License.

---

💡 Built with Flutter, powered by offline AI using TensorFlow Lite.