Design And Implementation Of Moving Object Detection And Tracking System

Video moving object detection and tracking technique, a new research subject brought forth by the development of digital video, has a wide range of application and a greatly prospects for development in many aspects such as national security, civilian career and so on.The main work of this paper is as follows:In moving target detection area, in order to acquire the motion target this paper combined background subtraction based on Gaussian mixture model and symmetric frame difference. For the obtained motion target contains some shadow pixels, improved standards for detection and removal of shadow are used. Compared with the original algorithm, the improved standards reduce the deal time for average each frame by18%to30%and still obtain batter results of removal of shadow. After the removal of shadow we gained the final precise motion target.In target tracking area, on the one hand as the creation of the target histogram in Camshift algorithms effects the precise of target tracking directly, so in this paper the creation method of the target histogram is improved and "weight histogram" is proposed to create target histogram and this new method significantly improved the presion of target tracking; on the other hand, with regard to the situation of tracking target failure which caused by the tracking target,because the trarget was occluded or the target moved too quickly, this paper combining Kalman filter and Camshift algorithm to predict and track the moving target. And with this method, it is greatly improved the robustness of the tracking algorithm.In this paper, moving object detection and tracking system is implemented by using MFC to build framework and combining OpenCV2.0in Visual Studio2008development environment ultimately. This system can rack moving targets with19frames per second averagely and it is basically meet the real-time tracking of moving target in low frame rate video.The implementation of shadow removing method in this paper can reduce the time of removing shadow after the moving target detection. This system is able to track the moving target which has distinct color features, so it can be used in the fields of hand tracking. Besides,the system can be extended to the intelligent robot vision tracking based on binocular stereo vision tracking,so it can be applied to the football, basketball, and fire-fight competition for intelligent robot.