Traffic management is a major challenge in cities where the roads are congested with a greater number of vehicles than their capacity. This congestion leads to delays, affects the productivity and also, contributes to air pollution. Current traffic control systems are static and ineffective in handling the complex traffic situations. Hence, the need of an automatic traffic control system that can effectively control and manage the real time traffic. The proposed system uses live images from traffic cameras to detect the vehicle type and calculate the traffic density with Image Processing and Artificial Intelligence. A specific model is taught to identify the various kinds of vehicles, such as cars, motorbikes, auto rickshaws, and larger vehicles, and the You Only Look Once (YOLO) method is utilized to detect the vehicles. Also, this system changes the traffic lights dependent on the current traffic density to lessen potential clogs, speed up transit, and clean up the air. OpenCV can draw coordinates on images using the object detection output, which is in JSON format. A cloud computing system calculates cycle time and updates signal timing based on traffic density. This approach has the potential to reduce vehicle waiting time, congestion, maintenance costs, and air pollution.
Automated Traffic Signal System Incorporating Real-Time Traffic
Lecture Notes in Civil Engineering
International Conference on Innovative Methods and Practical Applications for Cognizant Transportation Systems ; 2023 ; Thiruvananthapuram, India March 20, 2023 - March 22, 2023
Cognizant Transportation Systems: Challenges and Opportunities ; Chapter : 40 ; 531-541
2024-11-30
11 pages
Article/Chapter (Book)
Electronic Resource
English
TRAFFIC SIGNAL CONTROL INCORPORATING INFORMATION ON NON-MOTORED TRAFFIC
European Patent Office | 2018
|Online Stochastic Routing Incorporating Real-Time Traffic Information
Online Contents | 2013
|Online Stochastic Routing Incorporating Real-Time Traffic Information
Transportation Research Record | 2013
|European Patent Office | 2021
|