A Method Of Traffic Flow Statistics In Video Based On Background Difference Method And Virtual Region

Nowadays, the traffic flow statistics and detection has become an area of "intense research", as an important technology of intelligent transportation system. In the intelligent transportation system, traffic flow is an important parameter to analyze the transport situation, such as through the statistics of the vehicle flow to make the ease design and ensure the safety of bridge traffic, and design the traffic lights transform according to the traffic situation, so as to solve the traffic problem. Traffic information also can be used as an important parameter of other functions in the intelligent transportation system.In order to make the intelligent monitoring system can better play to its advantages, and increase the scalability of the intelligent monitoring system, this paper studies the method for video traffic statistics which based on integration of the background difference and the virtual regional. Mainly studies the applicable target detection techniques and related algorithms for traffic surveillance video in order to realize the traffic flow statistics. This paper studies the commonly used target detection algorithm, including five kinds of the edge detection algorithm, and the comparison of several target detection algorithm for the video vehicle detection. Also studied several commonly used methods of background modeling and analysis of their advantages and disadvantages.In this paper, Traffic statistics method is completed in a variety of city road environment.First, the background modeling is established by the mean background modeling method,then use the background difference method to detect moving objects.Then a virtual detection zone was set up in the appropriate position of the video image of the road.The identification and classification of moving targets is realized through connected image block area of the foreground in the virtual area.Finally, according to the passing vehicle number of the virtual area,we can get traffic information,so as to get the data of traffic flow in the road video. Through the study and the experiment, finally realizes the traffic flow statistics. At last, analyzing the missing and false detection in Statistics...