Research On Video-based Intelligent Traffic Control System

Traffic congestion is company with the fast growth of world economy, and it is serious in developed country.The problem caused by traffic congestion not only bring about inconvenience of our journey,but also cause the lose of economy.Related research has been launched for several decades,and many technology have been put into use for the research of intelligent traffic system. Probability ,Fuzzy controll, Digital image process, Nervous network and genetic algorithm have been put into use for the design of intelligent traffic system. Unfortunately,in our country the effect of traffic congestion is getting worse,and the related research is poor.Even now the traffic light base on fixed time algorithmstill stand up well on road in many small city or town.The details of stable western intelligent traffic system is not open.It has great significance of developing our own intelligent traffic system.But the difficulty is to promote the capability of traffic and to detect the usefull parameter.Illuminated by the existing achievement,I design an real-time algorithm which calculate the signal period by the length of watting vehicle which detected by digital camera.When it exist several similar phase position meanwhile,I adopt the predetermination of the length of the coming period to decide the sequence of travel.We use RBF nervous network forpredeterminate.This paper include experiment and it's result about all the methods of my reserch. The control algolithm of this paper effectively reduces congestion and calculate the signal duration on the basis of demand of each phase position.But it still can enhance the stability of detection and the rationality of signal timing algorithm.