Project

Overview

An AI-powered Automatic Number Plate Recognition (ANPR) system that uses computer vision and deep learning to detect and extract vehicle license plate information from images and video streams in real-time.

The system leverages OpenCV for image preprocessing and contour detection, combined with a trained deep learning model using TensorFlow/Keras for character recognition. It can handle various lighting conditions, plate angles, and different plate formats.

This project demonstrates practical application of AI/ML techniques in a real-world scenario — useful for parking systems, traffic monitoring, toll booths, and law enforcement applications.

Key Features:

1. Real-time vehicle plate detection from video/images
2. Deep learning model for character segmentation & recognition
3. OpenCV preprocessing pipeline (grayscale, thresholding, edge detection)
4. Handles multiple plate formats and angles
5. Confidence scoring for character predictions
6. Export detected plates to CSV/database
7. Works under various lighting and weather conditions