Track Net – AI-Powered Vehicle Individual Tracking through Networked CCTV Cameras

Abstract

The rapid urbanization and exponential growth in vehicular movement have made traffic management, road safety, and surveillance critical challenges for modern cities. Traditional Closed-Circuit Television (CCTV) systems primarily serve as passive monitoring tools, relying heavily on human supervision and often resulting in inefficiencies, missed violations, and delayed responses. To address these limitations, this paper introduces TrackNet, an AI-powered surveillance system designed to detect and track vehicles and individuals through networked CCTV cameras. Leveraging advanced deep learning techniques such as YOLOv8 for real-time detection, TrackNet integrates multiple modules including Automatic Number Plate Recognition (ANPR), helmet detection, seatbelt usage detection, and three-seater violation monitoring. The system transforms conventional surveillance into an intelligent, automated enforcement platform capable of proactive safety monitoring. Experimental evaluation demonstrates that TrackNet improves detection accuracy, reduces false positives, and ensures scalability across urban and institutional environments. By combining efficiency, automation, and real-time analytics, TrackNet contributes significantly to smart city initiatives and public safety enforcement.