TRI-GATE is an AI-powered vehicle gate access system designed to enhance security and efficiency. It leverages advanced machine learning and computer vision techniques to verify vehicles, license plates, and drivers in real-time, ensuring a seamless and secure access experience.
-
License Plate Recognition:
- Detects and authenticates license plates, including Saudi-specific formats.
-
Vehicle Classification:
- Identifies vehicle make and model (e.g., Toyota Camry, Honda Accord, Ford Crown Victoria).
-
Facial Recognition:
- Authenticates drivers using advanced facial recognition (InsightFace).
-
Real-Time Performance:
- Provides fast and reliable detection and authentication.
- Python3.9
- tensorflow==2.16.1
- torch==2.3.0
- onnx==1.16.0
- onnxruntime==1.17.3
- opencv-python==4.9.0.80
- scikit-image==0.23.2
- albumentations==1.4.6
- numpy==1.26.4
- scipy==1.13.0
- psycopg2-binary==2.9.9
- matplotlib==3.8.4
- insightface==0.7.3
- rich==13.7.1
- tqdm==4.66.4
- Captures video input or live stream to detect vehicles, license plates, and faces.
- Verifies the 3 features against a database.
- Opens the gate for verified entries.
InputVideo.mp4
-
Clone the Repository:
git clone https://github.com/QEC-Team/TRI-GATE.git
-
Navigate to the Project Directory:
cd TRI-GATE -
Install Dependencies:
pip install -r requirements.txt
-
License Plate Detection:
-
Gather your own dataset of images of cars that have license plates on them and put it into 'License_Plate_Detection_Dataset' folder
-
Label (rename) each image with the X and Y coordinates of the license plate, as shown in the image below.
-
-
Run 'MTCNN/data_set/preprocess.py' to split training data and validation data and put in "lpd_train" and "lpd_val" folders, respectively.
-
Run 'MTCNN/data_preprocessing/assemble_Pnet_imglist.py'
-
Run 'MTCNN/data_preprocessing/gen_Pnet_train_data.py'
-
Note: Repeat this process for Train and Validation.
-
Run 'MTCNN/train/Train_Pnet.py'
-
Run 'MTCNN/data_preprocessing/assemble_Onet_imglist.py'
-
Run 'MTCNN/data_preprocessing/get_Onet_train_data.py'
-
Note: Repeat this process for Train and Validation.
-
Run 'MTCNN/train/Train_Onet.py'
-
Collect pnet_Weights & onet_Weights and put them into the weights folder.
-
-
License Plate Recognition:
- Gather your own dataset of images of pure license plates and put them in folder 'License_Plate_Recognition_Dataset'.
- Image size must be 94X24.
- Rename each image to the license plate characters (e.g., ABC1234.jpg).
- Run 'LPRNet/data/preprocess.py' to prepare the dataset.
- Run 'LPRNet/LPRNet_Train.py' for training.
- Ensure PostgreSQL is installed and running.
- Create a database using SQL (see Access_database.sql file for reference)
psql -U <username> -c "CREATE DATABASE <database_name>;"- Execute the provided SQL file to initialize the database:
psql -U <username> -d <database_name> -f database.sql- `<username>`: Your PostgreSQL username.
- `<database_name>`: Desired database name.
python3 main.pyTo use our prepared training datasets:
These folders are empty in the repository and must be downloaded separately due to size limits.
Contributions are welcome! Follow these steps:
- Fork the repository.
- Create a new branch for your feature or bug fix.
- Submit a pull request with a detailed explanation.
For questions or suggestions, feel free to reach out to the project team:
- Abdulaziz Alorf: aaalorf@qu.edu.sa
- Muhannad Alsultan: 471114070@qu.edu.sa
- Bander Alwazzan: 471113811@qu.edu.sa
- Thamer Alghonaim: T.Alghonaim@signalinlab.ai
- Faisal Alasakakir: F.Alsakakir@signalinlab.ai
- Abdullah Alhassan: A.Alhassan@signalinlab.ai
This project is licensed under the MIT License.
See the LICENSE file for more details.