AI-BASED REAL-TIME AUTOMATIC NUMBER PLATE RECOGNITION SYSTEM

Abstract

The increasing density of vehicles in modern transportation systems has created a critical need for automated and intelligent monitoring solutions. Vehicle number plate detection and recognition systems have become essential for applications such as traffic regulation, surveillance, toll automation, and identification of unauthorized or stolen vehicles. This research presents an advanced real-time vehicle number plate recognition framework that integrates computer vision and deep learning techniques for efficient performance.

The proposed system processes live video streams or image inputs to detect moving vehicles and accurately localize their number plates using a deep learning–based object detection model. The extracted plate regions are further processed through Optical Character Recognition (OCR) to convert visual text into digital format. Image preprocessing techniques, including noise reduction, contrast enhancement, and grayscale conversion, are employed to improve detection accuracy under varying environmental conditions such as poor lighting, motion blur, and different viewing angles.

To enhance reliability, the system combines a YOLO-based detection approach with Tesseract OCR for robust character recognition. The recognized license plate data is systematically stored in a database and automatically verified against a predefined list of flagged or unauthorized vehicles.

The developed framework demonstrates high accuracy, real-time capability, and scalability, making it suitable for intelligent transportation systems. It offers a cost-effective and automated solution for vehicle identification, contributing to improved road safety, efficient traffic management, and enhanced security monitoring.