Smart Traffic Light System Time Prediction Using Binary Images
Annapureddy Bhavani, Sandeep Verma, S. Vikram Singh, Suman Avdhesh Yadav
Abstract
In current circumstances, congestion in traffic is a major issue. Due to traffic congestion, many people didn’t reach their destination place in a pre-defined time. This may lead to loss of job opportunity, miss of journey, etc. Many algorithms had been developed for reducing the congestion problem in traffic. They are working well in day times but those algorithms didn’t give precise results at night times. So a new algorithm is developed by using image processing in this paper that will give precise results at night times also. This is done by isolating the RGB image into its individual channels and converting each channel to binary images with different thresholds for vehicle identification. Based on the results, the density of traffic is estimated and compared with the reference one with no traffic to get the allowed time for the green light to glow. So the allowed time is completely dependent on the density of the traffic.