Research Of The Moving Video Target Detection And Tracking Arithmetic

With the development of the science and technology, the research of using computer to imitate eyes to identify objects is more and more important. Among the applications, the arithmetic of the motion detection and tracking is the core. In this paper, a new arithmetic of the motion detection and tracking is presented.It can be divided into the following parts. Firstly, an ideal background model is obtained by a statistical method from the video data acquired by the stationary cinematograph. Secondly, the model is real-time updated by the new background data to make it much more robust. Thirdly, the motion objects are detected by background subtraction method. Fourthly, the shadows of the motion objects are detected in the HSV, and a back-processing method of filtering noises is adopted to make the results more satisfactory. Finally, matching the detected motion objects with the minimal outer rectangle realize the tracking. The experimental results are satisfactory.