A Computer Vision-Based System for Real-Time Vehicle Speed Detection and Automated Complaint Logging

Vehicular speeding, particularly in residential and school zones, is a major public safety concern. Traditional speed enforcement methods are often resource-intensive and provide only sporadic coverage. This paper presents a low-cost, automated system that leverages computer vision to monitor traffic, detect speeding vehicles, and log complaints in real-time. The system processes a video feed from a standard camera to perform vehicle detection and tracking using the YOLO (You Only Look Once) and Deep SORT algorithms, respectively. By performing a perspective transformation via camera calibration, the system accurately maps pixel coordinates to real-world distances, enabling robust speed estimation. When a vehicle is detected exceeding a user-defined speed limit, the system automatically logs a “complaint” or incident report, which includes the timestamp, estimated speed, and video evidence, into a database. This data is made accessible through a web-based interface. This framework offers an accessible and data-driven tool for communities and local authorities to monitor traffic safety, identify speeding hotspots, and justify the need for traffic calming measures.