This project aims to track unauthorized vehicles entering a society from the CCTV camera footages using YOLO-V5 model to detect the license plates. The model was trained on Car plate detection dataset. Then the license plate numbers were read from the license plates using EasyOCR and they were entered in the database

Authorized residents and vehicles can be registered by using the UI made on the flask app

1. Clone the github repository

git clone https://github.com/aayush1036/vehicle-security.git

2. Create virtual environment

virtualenv env 
# For linux 
source env/bin/activate
# For windows 
env\Scripts\activate

3. Install the required dependencies

pip install -r requirements.txt

4. Edit the config file to suit your database credentials

5. Start the detections by using the detect.py script on sample video or your CCTV/webcam

# For detecting on sample video
python detect.py --society YOUR_SOCIETY_NAME --position ENTRY/EXIT --path sample.mp4
# For detecting on CCTV/Webcam
python detect.py --society YOUR_SOCIETY_NAME --position ENTRY/EXIT --device 0/1/2....

6. If you want to retrain the model on custom data, please refer to the kaggle notebook published on License plate detection notebook by me and please upvote it and comment on it if you liked it

1. Home page
   

Detailed explanation of the project can be found on Medium Article. Please upvote the article if you find it helpful

1. Improve OCR
2. Make predictions more faster