Soccer Ball Tracking using Dynamic Kalman Filter with Velocity Control.pdf

2009 Sixth International Conference on Computer Graphics, Imaging and Visualization Soccer Ball Tracking using Dynamic Kalman Filter with Velocity Control Jong-Yun Kim, Tae-Yong Kim GSAIM, Chung-Ang Univ South Korea { idsrge@, kimty@cau.ac.kr } Abstract which team has superiority in the game and also we can In this paper, we propose the ball tracking method track the viewers ROI (Region of Interest) part by that is tracking the ball adaptively and robustly in the chasing a ball position in the soccer video. But it is very soccer video. In the latest works, people have used the difficult to detect and track the ball in the soccer video Typical Kalman Filter to track the ball. But when the because the size of the ball is too small and it has only a ball is disappearing due to the occlusion with players, little amount of features. So many researchers study the Typical Kalman Filter has no choice but to make a pool method to solve these problems recently. prediction and especially if the player take the ball for a In the past time, people used the feature of long time, the error is produced much more. To appearance of the object, seems like a color, shape, size, overcome these problems, we propose the Dynamic texture to track the ball in the video. T. DOrazio used Kalman Filter algorithm. Dynamic