This project do about vietnam license plates recognition, we use a self-build dataset for training detect & recognition model
At License plate detect we use YOLOv4
At Text detection on License plate we use CRAFT
At Text recognition on License plate we use Star-Net
# Clone this project
$ git clone https://github.com/hieu28022000/LP_Reg.git
# Access
$ cd LP_reg
# Install dependencies
$ pip install -r requirements.txt
# download pretrain model
./download_model.sh
# unzip pretrain models
unzip models.zip
# run pipeline on images folder
python3 pipeline.py --run_on_folder True --folder_path <path to folder image>
# or run pipeline on image
python3 pipeline.py --run_on_folder False --image_path <path to image>
You can see full project and result in here