Urban traffic congestion has become a major challenge due to the rapid increase in the number of vehicles on roads. Conventional traffic signal systems operate based on fixed time intervals and do not adapt to real-time traffic conditions, leading to inefficient traffic flow and unnecessary delays. This paper proposes an AI-based adaptive traffic signal control system that dynamically adjusts signal timings based on real-time traffic density. The proposed system utilizes cameras installed at road intersections to capture live traffic footage. The captured images are processed using techniques from Computer Vision and Artificial Intelligence to detect and count vehicles present in each lane. Based on the detected traffic density, the system dynamically allocates green signal duration for different lanes, prioritizing lanes with higher vehicle density. The implementation uses object detection models trained on the COCO Dataset to identify vehicles such as cars, buses, trucks, and motorcycles. By analyzing traffic conditions in real time, the system optimizes signal timing to reduce congestion, waiting time, and fuel consumption. Experimental results demonstrate that the proposed system improves traffic efficiency compared to traditional fixed-timer traffic signal systems. The proposed solution contributes toward the development of intelligent transportation systems and supports smart city initiatives.
Ms. Renuga R M.E(PhD), Katheejathunnisa A, Megashree P R, Yuvarani G Department of Computer Science & Engineering Velammal Institute of Technology, Panchetti (Wed,) studied this question.