| Since the electronic computer came out, intelligent tools improve the standard of living and the production efficiency, the traditional work done by hand has gradually been replaced by it. In recent years, due to the considerations of the security of people and public and other, the detection of moving targets in dynamic video sequence has played a very important roles on video surveillance system,traffic automatic monitoring, target detection and tracking, target recognition and so on.the motility of the target and the influetion of the surrounding environment make the algorithm complex. The extraction of target in video cameras is realized by camera, because of the mutility of the camera, the algorithm of the target tracking has two direction, one is the target tracking in static background, the other is the target tracking in dynamic background. There are great differences in two algorithm. In first one, the only change is the target, so in this condition, the background is changed hardly, the algorithm is based on calculus of variations;in the second condition, the target tracking is completed in the condition of camera moving. As the motility of camera lead to the vary of background, the algorithm of first condition would lose the effect, so we need find a new algorithm to achieve target tracking.In dynamic background, the algorithm of target tracking is most based on the algorithm of Mean Shift, the first part of this paper introduce the devrlopment of the algorithm of target tracking, to give a theoretical basis and the research background for the content of this paper.In the second part, this paper talks about the main theory of Mean Shift, which is the most basic theory, so as to comprehend the main idea of target tracking from the standpoint of math.Since the basic method could not satisfy the accuracy of object tracking fully, so by combining Kalman calculate with Mean Shift, we can raise the accuracy, so the third part of the paper present the method of Kalman prediction.In the fourth part of this paper, you will see some experiment, those experiment will give the result of the MS and the Kalman prediction, then be aim at the disadvantage of the method refered, a new method is propsed.At the same time, the experiments also will show the advantage of this method.At the end of the paper, the author makes a brief summarization and points out the promising direction for future research. |
