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.