Kernel Based Object Tracking Using Mean Shift Method

摘要

In this age of dramatic technology shift, one of the most significant development has been the emergence of digital video as an important aspect of daily life. While the Internet has significantly changed the way in which we obtain the information, it is much more attractive because of the powerful medium of video. In this paper we have described kernel based object tracking algorithm using mean shift method. The goal of an object tracking algorithm is to generate the trajectory of an object over time by locating its position in every frame of the video. There are various applications of object tracking in the field of computer vision. A smart camera is a very important component for many applications such as, video surveillance, traffic monitoring system and for mobile robots.