Moving Object Recognition Based On OpenCV

With the increasingly high performance video surveillance equipment and the price become increasingly popular, the associated security technology has also been extensive research and application, especially in the foreground detection algorithm, the moving target tracking algorithm, the moving target recognition has become related research in the field of hot spots, has a very broad application market.The research objectives of this paper the sensitive direction detection, the future detection, moving target tracking, moving target trajectory handling algorithm in detail. The background of rapid change, the prospects for multi-Gaussian detection model and the goal of removing the algorithm as a moving target detection, the method is effective to reduce the background noise on the foreground image extracted, focused on the theoretical basis of the CamShift tracking algorithm Kalman algorithm and program implementation process. Finally OpenCV visual software and MFC for secondary development, and ultimately the design of the software, moving target detection, a new target detection, object tracking, sensitive to the direction of recognition of the moving target, moving target video segment extraction.Finally, using this software has been extensively tested and the results were summarized and analyzed the sensitive direction, can be drawn by analyzing the test data:the sensitive direction detection software designed in this paper can effectively track moving objects, the specific direction of moving objects can effectively extract. This software on VC platform, the next step, the most important job is how to effectively transplant work and further improve the recognition efficiency of the software.