City Traffic Congestion Degree Based On Video Analysis Technology

With the development of society and improvement of living standards, increase in vehicles poses a serious threat to road traffic, but more importantly, road traffic congestion problems have seriously affected the transportation of citizens, and restricted social and economic development. Growth speed of vehicles is too fast. The current transport network can not meet the huge amount of traffic. Although the Government has introduced a series of traffic laws and regulations, but still can not fundamentally solve the traffic problem in most cities in China, especially in large cities serious traffic congestion problems still exist.This paper describes the analysis techniques. The text includes three parts, namely vehicle detection, vehicle tracking and the determining of traffic congestion level. The purpose of this article is to determine the traffic congestion level based on such traffic parameters as space occupancy rate and speed.The paper analyses the congestion according to video-based vehicle detection and in-depth study on the tracked vehicles.The research work is as follows:(1) Analysis of the existing detection methods, including background subtraction, frame difference method, optical flow. In this paper, vehicle detection is done with an improved single-Gaussian model of background difference,(2)Due to the shortcomings of existing tracking methods, Harris corner points of multi-scale and mean-shift vehicle tracking method is applied. Kalman filter prediction method is used. This paper is the contrast of the image before and after the extraction is completed through the corner of the vehicle.(3) According to vehicle speed and space occupancy as the traffic parameters of the analysis of vehicle traffic congestion extent, there are four degrees:smooth, slight congestion, congestion, serious congestion. It supported vector regression as a traffic forecasting model. This paper conducted a simulation to predict the degree of congestion on the road.The proposed algorithm and technology programs in this paper are tested and further proved to be viable through experiments...