Detection And Tracking Of Moving Vehicle In Video Sequence Based On OpenCV

Urban traffic congestion is a common problem faced by all governments in the world, the development of intelligent transportation system can effectively improve the traffic situation. As an important part of intelligent transportation system, vehicle detection and tracking has become a hot research topic. Vehicle detection and tracking not only need a lot of data processing for video signal, but also make the corresponding judgment to the result, such as vehicle behavior analysis, motion track and so on, all of these require accurate algorithms. The compilation of a large number of underlying code is time consuming and power consuming, the Intel company's open source code library Open CV, its function library contains a wealth of image processing source code, provide a great help for the smooth progress of the study. Based on the Open CV two development platform, in this thesis mainly do the image preprocessing, detection and tracking of moving vehicles in the video sequence,and the design of system software three aspects of the work.In the preprocessing part, aiming at the influence of the factors such as noise and color similarity, which are easy to appear in the video frame sequence image, image denoising and morphological processing, for the image of the target feature information is more obvious, is conducive to the detection and tracking of the target in the following thesis.In the moving vehicle detection section, based on the traditional detection algorithm for moving vehicle detection result is not accurate. In this thesis, the three frame difference method and edge information are combined to detect the vehicle. Firstly edge detection operator is to detect the edge information of the vehicle, then the three frame difference method is used to make the difference between three consecutive frames. The operation logical-AND for the results of the differential operation, then morphological processing. Finally, the better detection results are obtained, and the feasibility of the algorithm is verified by experiments.In the moving vehicle tracking section, taking into account the research of this thesis is the camera to take the road video sequence image, this environment is easy to be blocked by the vehicle, the target color and environmental color similarity and other factors of interference. So this thesis uses the Kalman filter and Cam Shift algorithm to realize the tracking of moving vehicles, the feasibility of the algorithm is verified by the tracking of single vehicle and multi vehicle.Finally, the system of in video sequences moving vehicles detection and tracking is built in VS2010 and Open CV environment. The control interface of the system is mainly composed of three parts, the control area, the video information and the video broadcast area, can be achieved on the two values of the video sequence image, vehicle edge detection and tracking, and other basic functions. And the effect of the system on the detection and tracking of moving vehicles is tested and verified.