Automatic Number Plate Recognition

Automatic Number Plate Recognition (ANPR) is a computer vision project designed to detect and recognize vehicle number plates from images and video streams. This repository provides an end-to-end solution for extracting license plate information using image processing and deep learning techniques.

Features

• License Plate Detection: Locates number plates within images or video frames.
• Character Segmentation: Segments individual characters from detected plates.
• Character Recognition: Recognizes alphanumeric characters using trained models.
• End-to-End Pipeline: Integrates detection, segmentation, and recognition in a streamlined workflow.
• Extensible Design: Easily adaptable for various regions and plate formats.

Technologies Used

• Programming Language: Python
• Libraries: OpenCV, NumPy, TensorFlow/Keras or PyTorch, Tesseract OCR, Matplotlib
• Deep Learning: CNN-based models for detection and recognition

Installation

1. Clone the Repository

   git clone https://github.com/bnaveenbharathi/Automatic-Number-Plate-Recognition.git
   cd Automatic-Number-Plate-Recognition

2. Install Dependencies

   pip install -r requirements.txt

   Additional dependencies:

   • Tesseract OCR (if using Tesseract)
     • Install via: sudo apt-get install tesseract-ocr (Linux) or download from here

Usage

You can use the ANPR pipeline on images or video files.

Run on an Image

python app.py --image path/to/image.jpg

Run on a Video

python app.py --video path/to/video.mp4

Example Output

The program will output the detected number plate and recognized text in the terminal and optionally display or save the results.

How It Works

1. Detection: Uses image processing and/or deep learning (YOLO, Haar cascades, etc.) to locate the license plate region.
2. Segmentation: Extracts individual characters from the detected plate using contour analysis or deep learning.
3. Recognition: Identifies each character using an OCR engine (Tesseract) or a trained CNN classifier.
4. Display/Export: Outputs the recognized plate number and annotated image/video.

Training

• Data: Use labeled images of license plates and characters for training detection and recognition models.
• Model: Train custom CNN models for character recognition or use pre-trained models (e.g., Tesseract).

Contributing

Feel free to fork the repository and submit pull requests for improvements or bug fixes.

1. Fork the repo
2. Create a new branch (git checkout -b feature-name)
3. Make your changes
4. Commit and push (git commit -am 'Add new feature')
5. Open a pull request

License

This project is licensed under the MIT License.

Acknowledgements

• OpenCV team for image processing tools
• Tesseract OCR
• Dataset contributors

Contact

For questions or support, open an issue or contact bnaveenbharathi.