| Abstract: |
Traffic congestion and public safety remain two of the most pressing concerns in today’s fast-growing cities. Long queues at traffic signals waste time and fuel, increase pollution, and frustrate commuters, while security threats demand quicker and more reliable law enforcement responses. This paper introduces a practical, low-cost solution that tackles both problems together—an IoT and OpenCV-based Smart City Traffic Management and Vigilante Detection System. Our system uses infrared (IR) sensors to continuously monitor traffic density on each lane. Instead of relying on rigid signal timers, it adjusts green light durations in real time—5 seconds for moderate traffic and 8 seconds for heavier congestion—guided by a priority-based algorithm. If traffic builds up on multiple lanes, predefined rules decide the sequence, and in the event of prolonged jams, the system automatically issues rerouting alerts to ease the bottleneck. On the security front, a facial recognition module built with OpenCV, Haar-Cascade classifiers, and the Face Recognition library scans live video from surveillance cameras, comparing detected faces against a stored criminal database. If a match is found, the system instantly retrieves the person’s details, pinpoints their GPS location via the GEOPY library, and emails an alert—with photo and information—to law enforcement. Developed using an Arduino Uno, Python-based processing, and a wireless camera, the system demonstrated perfect accuracy in IR-based vehicle detection, reduced average wait times by 14% over fixed-timing signals, and reliably triggered criminal alerts during testing. By combining smart traffic control with real-time criminal detection in a single platform, this approach offers a scalable blueprint for safer, more efficient cities. Future upgrades aim to incorporate AI-driven traffic prediction, deep learning-based face recognition, and cloud-based analytics for seamless, city-wide operation. [ABSTRACT FROM AUTHOR] |