Published Paper Details:

This project tackles the problem of traffic congestion at crossroads and problems faced by traffic congestion at crossroads. Our idea is to create system which will count the number of vehicles in each lane of crossroads and by analyzing the count green light time and red-light time will be adjusted. i.e., the lane containing less vehicles will get less green time than usual and time reduced form such lanes will be added to the green light time of lanes containing maximum number of vehicles Existing system uses costly density sensors but we are going to use cameras installed on cross roads which will give inputs to our YOLO5 algorithm for counting vehicles and then input will be analyzed by machine learning model to adjust the timings of signal. The movement of cars through the intersection of numerous roads is monitored and managed by traffic signal control systems. Their goal is to make cars move smoothly in the routes for transportation. But because there are so many variables to consider, coordinating several traffic signal systems at nearby crossings is a challenging task. Approaching junctions with changing flows are not handled by conventional methods. Furthermore, the current traffic system does not account for accidents, emergency vehicle passing, pedestrian crossings, conflicting traffic light systems, and differences in vehicle flow over time. This results in congestion and traffic jams.