The Research Of Mean Shift Object Tracking Algorithm

Object Tracking has always been a hot issue in Computer Version research,its application area include Video Surveillance,Human-Machine,Virtual Reality and so on.It requires accurate and effective track in different background,light changes, object block and so on.Track the target for stable is a hot research issue in the filed of object tracking.The Mean Shift tracking algorithm basing on the Mean Shift use the color histogram to track object, because of its simpleness, perfect real-time performance,the ability to process the target deformation and the block situation and other difficult situation,and also the high accuracy, so the Mean Shift video tracking algorithm has been widely used.The traditional Mean Shift tracking algorithm uses the color histogram to track object.The color-histogram has some excellences-the rotation inflexibility and scale inflexibility and so on, but the color histogram does not involve the spatial distributing information of the color. In this paper, we propose a spatial color histogram,based on this point we propose a spatial color histogram Mean Shift video tracking algorithm.The experiment shows that this method can track the target object better than the classic Mean Shift tracking Algorithm while meeting the real-time requiring.Mean Shift Tracking with Self-Updating Tracking Window, it use the relation between the information and the scale of the image to adjust the tracking window automatically. But the algorithm ignore the Character Point(CP)'s location when calculate the information of the image.The CP more close to the center of the tracking window more strong it can reveal the character of the tracking target, On the contrary the weak, so the CP close to the window edge and the CP close to the window center could not be in the same position.This paper improve the algorithm by add a kernel function as a weight function,so it can lower the edge CP affect to the information. Experimental result demonstrated that the improved algorithm could select the proper size of the tracking window.