An intelligent video analytics-based traffic lights system.

Abstract

This paper presents an innovative approach to traffic management through the development of an intelligent video analytics-based traffic lights system. With the ever-increasing volume of vehicles on roads, conventional traffic control methods struggle to efficiently manage traffic flow and reduce congestion. Our system leverages video analytics techniques to analyze real-time video data captured from traffic cameras. By applying object detection, tracking, and classification algorithms, the system can accurately identify, count, and track vehicles and pedestrians on the scene. This rich information is then utilized by the decision-making logic to optimize traffic light timings and facilitate the smooth flow of vehicles through junctions based on real-time demands. The proposed system offers several advantages over traditional traffic light systems, including adaptability to dynamic traffic conditions in real-time. Experimental testing results showed 99.996% of vehicle and pedestrian detection with real time response by the decision-making logic which demonstrates the effectiveness of the system in improving efficiency during times of no jams and equally attending to all lanes during jams thus reducing congestion and improving overall road safety. The system also provides for pedestrians crossing when it is safe and convenient. This research contributes to the growing field of intelligent transportation systems and offers a promising solution for smarter traffic management in urban areas.