A Survey on IntelligentTraffic Control System Using Image Processing

Abstract

The present traffic control systems depends upon the manual operation which leads to unnecessary wastage of fuel and time, which again leads to increase in the traffic congestion. More time is wasted at the signals even if there are no vehicles on the lane because the signals have fixed timing for all roads which is not dependent of traffic queue. Some nonuniform illuminations that come from artificial light sources affect the night-time surveillance. Artificial light sources result in glow that the objects or the vehicles present near the light sources will not be visible. Web camera is fixed and the images of all the four lanes will be captured. The method used in this paper removes the effect of glow through image gradient decomposition. The enhanced image is obtained through Poisson solver. The real time vehicle density is measured to control the traffic system using canny edge detection algorithm and the vehicle count is calculated for each path or road and comparison is made with the other roads. The green signal and red signal are received by the roads with the maximum vehicles and minimum vehicles respectively.