Research On Intelligent Traffic Light Control System Based On Image Processing

With the rapid development of China’s economy,the people’s living standards have been greatly improved.In terms of travel,China’s car ownership has increased rapidly,but too many cars have made the city’s traffic conditions very congested.Traffic congestion frequently occurs at the traffic intersections of cities.At present,most of the traffic lights in China are controlled by means of timed switching.They are not intelligent and cannot be automatically adjusted and switched according to the real-time conditions of intersections,making traffic passages less efficient.It is easy to cause traffic congestion.This thesis analyzes the vehicle information at the traffic intersection based on the image processing method,and then controls the traffic lights at the traffic intersection in real time,mainly including the following work:First,extraction of vehicle prospects at urban traffic intersections.This thesis uses the optimized algorithm based on the highest frequency background modeling to model the background of the intersection.Then the background difference method is used to extract the information of the vehicle foreground.The threshold image segmentation method is used to extract the vehicle information of the foreground image.The image is morphologically processed to obtain an effective foreground vehicle target,and finally the linear equation method is used to segment the lane portion of the effective detection area.Second,design an algorithm for detecting the queue length and traffic flow of intersection vehicles.For the detection of the queue length of the vehicle,the sliding window method can be used to effectively detect the approximate position of the vehicle team,and then the sliding line method is proposed based on the sliding window method for accurately detecting the head and tail positions of the vehicle team.Then,using the fitting method,the transformation relationship between the image pixel point and the actual three-dimensional space distance point is obtained,and the actual length of the vehicle team is obtained according to the obtained transformation relationship.The vehicle traffic detection algorithm is to set a detection area at the intersection,and determine the number of vehicles passing according to the pixel change in the detection area.Third,design of Fuzzy Control System Based on queue length and two traffic parameters.Fuzzy control system herein is designed as a traffic lane and a queue length two input systems,the phase of the transit time of the green light output of the system.Compared with the effect of traditional timing control,the fuzzy control algorithm can effectively improve the traffic capacity of traffic intersections.