Auto License Plate Reader

Overview

Auto License Plate Reader (ALPR) is a computer vision project that identifies license plates and recognizes their characters. This system is built using YOLOv8 for object detection and supports both webcam and video file inputs. The goal is to create an efficient and accurate license plate recognition system that can handle multiple plate formats.

Features

• License Plate Detection: Detects license plates on vehicles.
• Character Recognition: Recognizes and extracts text from detected license plates.
• Multiple Input Sources: Supports real-time webcam input and video file processing.
• SqlLite DB: Stores real time numbers to a database
• Live Notifications: Send realtime notifications via telegram

Installation

Clone the Repository

git clone https://github.com/donsolo-khalifa/autoLicensePlateReader.git
cd autoLicensePlateReader

Install Dependencies

pip install -r requirements.txt

Model Files

Add your license plate model to the autoLicensePlateReader directory and rename it to best.pt. You can download my pretrained model here.

Video File

Ensure the video is present in the autoLicensePlateReader directory and name it licensevid.mp4 or change the following line in main.py to switch to webcam mode:

cap = cv2.VideoCapture("licensevid.mp4")  # Change to cap = cv2.VideoCapture(0) for webcam

Usage

Running the Detection

• Create an empty folder called json in the autoLicensePlateReader
• Run the following commands:

python sqlLiteDB.py
python main.py

Telegram Bot

To enable live notifications:

1. Set up a Telegram bot using BotFather.
2. Uncomment the following lines in main.py (lines 122 and 123):

message = "🚘 Detected License Plates:\n" + plate_text
threading.Thread(target=sendPlate, args=(message,)).start()

3. Create a .env file in the autoLicensePlateReader directory and add your token and chat ID:

TOKEN = add token here
CHAT_ID = add chat id here

TODO

• Improve OCR accuracy for better character recognition.
• Enhance multi-format license plate support.
• Optimize detection speed for real-time performance.

Contributing

Feel free to fork the repository and submit pull requests. Issues and suggestions are welcome!

License

This project is open-source.

Acknowledgments

• Ultralytics YOLO
• OpenCV Community
• PyTorch Developers
• Sort tracking algorithm