| In recent years, with the number of vehicles increasing, the problems of such as traffic congestion and so on become more serious.The status of Intelligent Transportation System (ITS) is rising in traffic management.Traffic flow analysis is one of the key of ITS and is the premise of intelligent transportation induced and control. There are many traffic detection methods, including computer vision technology based on image processing. It is becoming a hot spot of research in the ITS field, with intuitive and convenient analysis methods, counting accuracy, easy installation, and maintenance cost advantages gradually replaced the other methods, to obtain a wider range of applications.In the paper, we use a static single hole video camera to get the traffic scene as the research object, to extract scene traffic in the parameters for the purpose, with reference to the related literature at home and abroad, on the basis of several involved moving object detection method has summarized and comparison, analyzes the basic principle of vehicle detection methods, application and advantages and disadvantages, using a suitable for traffic parameter examination detecting method that implements a video traffic detection system. The main contents include:(1) Introduces three kinds of common moving object detection methods, and mainly describes the background of reduction algorithm is adopted in this paper.(2) Introduces several background modeling methods, considering the real time and detection algorithm of integrity, established in RGB color space is adopted in value method extracting method of background.(3) Realize a system based on video road traffic detecion, by setting a virtual inspection, extraction of vehicles on line, for information method of traffic detection.Finally, in this paper I programme simulation to the proposed algorithm on MATLAB 2007a in Windows operating platform.This algorithm based on the actual traffic flow image sequences were tested, and good results are obtained. |
PDF Full Text Request