Traffic Management using Computer Vision and SUMO

Traffic signals play a crucial role in the transportation system. They guarantee a safe drive at the road intersections. Traffic signals reduces traffic flow as the traditional signal scheduling gives green light to each direction for a predetermined time. In traditional signal scheduling, Green signal is allocated to all directions, one after another. If there is heavy traffic in one direction and less in another, Traditional scheduling method does not show any bias among them. This results in over traffic accumulation and traffic congestion at the road intersections. To give priority to the directions that has the maximum vehicle count, we propose an algorithm. The proposed algorithm takes vehicle count of all directions into consideration and gives immediate attention to direction where traffic is getting accumulated. The signal allocation does not only consider vehicle count. It is done in such a way that no direction gets to wait too long. Hence, the algorithm we propose gives equal importance to both Traffic density and Waiting Time. Moreover, it aims to minimize total waiting time of all the vehicles. We performed various experiments, from the obtained results, we were able to highly reduce the overall waiting time of vehicles in all directions.