ARCHIVES

With the rapid growth of urban transportation, vehicle theft has become a major public safety concern, creating significant challenges for law enforcement agencies and the general public. Traditional surveillance systems and manual monitoring methods are inefficient, costly, and incapable of providing real-time detection. This project proposes an AI-powered framework for stolen vehicle identification that integrates You Only Look Once (YOLOv8) for license plate detection and EasyOCR for text recognition. The system verifies recognized license plate numbers against a precompiled database of stolen vehicles and enhances situational awareness by logging geolocation data using IP-based detection. A Streamlit-based web interface is implemented to provide users with an interactive and responsive platform for image uploads, detection results, and real-time alerts. Experimental validation highlights the feasibility and effectiveness of combining deep learning–based object detection with optical character recognition in delivering scalable, low-cost, and user-friendly surveillance solutions. The modular design further supports future extensions such as live video stream analysis, API integration with national law enforcement databases, and automated notification services. Overall, this work establishes a practical, real-time, and cost-effective solution for modern vehicle surveillance and theft prevention.