The Research And Implementation Of Moving Object Detecting And Tracking In Video Monitor System

Moving object tracking is an important part of video monitor system, so how to track moving object steady, real-time and effectual ,becomes an important question that needs to be paid attention and researched.According to the questions, which include tracked object always charges pose, part been kept out and object circumrotates, the dissertation chooses an arithmetic( Meanshift) bases on object feature of color histogram that used to track moving object, which bases on lucubrating and analysing of object tracking system and comparing of arithmetic of objects tracking. According to fast constringency the arithmetic bases on color histogram constructs a similarity degree function that uses Meanshift arithmetic to search best matching, and tracks objects effectual.According to Meanshift arithmetic can not track fast object reliable, the dissertation imports Kalman Filter, and it is good to track fast object. Kalman Filter includes two models, signal model and observation model. It forecasts status vector in current frame according to observation vector in former frame, and finds the possible position of object. Then, it constitutes observation vector in current frame through the information of position, and forecasts status vector in next frame. It can substitute circulatory.The result of experiment indicate, the arithmetic (Meanshift) bases on object feature of color histogram can track moving object effectual in video monitor system, it keeps stability and continuity of moving object tracking, which has good value of the theoretical study and the broad market prospect.