Artificial Intelligence based Framework for Effective Performance of Traffic Light Control System

The rapid growth of innovations in all fields of science has made our lives easier, but the increase in traffic accidents on roads over the years has cost many lives. Local governments are unable to control the global economic growth that is accompanied by an increase in the number of automobiles on...

Full description

Saved in:
Bibliographic Details
Published in2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES) pp. 1 - 4
Main Authors Singh, Anuj, Sharma, Sachin, Purohit, Kamlesh Chandra, Nithin Kumar, K C
Format Conference Proceeding
LanguageEnglish
Published IEEE 24.09.2021
Subjects
Artificial Intelligence
Biological system modeling
DMM
Gray-scale
Local government
Roads
Technological innovation
Traffic
Traffic control
Traffic light auto control
Transportation
YOLO model
Online AccessGet full text

Cover

More Information
Summary:The rapid growth of innovations in all fields of science has made our lives easier, but the increase in traffic accidents on roads over the years has cost many lives. Local governments are unable to control the global economic growth that is accompanied by an increase in the number of automobiles on the road. Controlling traffic has been a problem for more than a decade and will continue to be a major concern in the near future. Despite the fact that numerous researchers presented their research findings, the problem remains unresolved. This work focuses on a novel approach to automated real-time traffic control based on artificial intelligence concepts. The videos were shot at a four-lane traffic signal in Dehradun and are being tested for various models capable of detecting and counting all types of vehicles. This research focuses on the development of a model that can automatically control traffic based on the YOLO model and DMM to control the traffic light. The YOLO model is integrated in such a way that traffic-related obstacles are minimized. The videos are taken with a 13mega pixel Camera in three places: morning, afternoon and evening. The gray-scale image subtraction system is used. The highest accuracy of the vehicle count is at a mean visibility of 96.15% in the morning, while the lowest accuracy of the fog/low visibility in the night is 66.66% It is also used to control traffic light automatically with the intelligence transportation system.
DOI:10.1109/ICSES52305.2021.9633913