Research On Intelligent Traffic Congestion Scheduling System Based On Machine Vision

With the increasing number of population and vehicles in today’s society,higher requirements are put forward for the transportation industry.The future of the city belongs to the era of intelligent transportation.However,there are a series of problems in the current urban traffic system,such as traffic congestion,traffic accidents,parking difficulties and so on.Among them,traffic congestion seriously affects people’s happiness,so this paper has carried out a series of analysis and research on the traffic congestion problem.In this paper,the machine vision technology is used to judge and grade traffic congestion,and the traffic congestion is alleviated by the way of intelligent timing of signal lights and vehicle traffic scheduling.Finally,the simulation experiment is carried out on the traffic situation at intersections.The main contents are as follows:First of all,the analysis of traffic congestion requires the acquisition of traffic indicators,the application of machine vision technology,the installation of cameras on the road,and the processing of the video shot.The number of vehicles was obtained by the background-difference method,and the background model which could be updated in real time was established by the mixed Gaussian background modeling,so as to avoid the influence of illumination on vision technology.Taking the virtual coil method as the theoretical basis,the virtual coil is constructed to count the vehicle travel and time and obtain the vehicle speed index.Secondly,it aims at the problem of traffic congestion judgment without fixed standard.According to DS evidence theory,the identification framework of traffic congestion is established.According to the number and speed of vehicles,a congestion classification model is built to identify different levels of congestion scenarios and adopt different countermeasures for different congestion scenarios.Finally,in view of the problem of how to relieve traffic congestion,the use of signal lights intelligent timing scheduling vehicle traffic.The working mode of the existing fixed timing signal lamp is improved.With the number and speed of vehicles as the input and the delay time of green light as the output,a fuzzy controller is built to realize the intelligent timing of the signal light to dispatch vehicle traffic.Using traffic simulation software as a platform,a comparative experiment was designed between the intersection with fixed distribution and the intersection with intelligent distribution.It was verified that the vehicle with intelligent timing of signal lights could improve the capacity of the road and alleviate traffic congestion.