Football players and ball trajectories projection from single camera's image

摘要

In this paper, we propose a method to track multiple players in a football match video which is captured by a single camera. The camera pans and brings the players and the ball into the view, enabling to record the whole pitch. The players' trajectories in frames and camera movement are obtained to estimate the trajectories on the pitch. Moreover, we propose an effective tracking method for solving occlusion problem by combining Particle Filter and Real AdaBoost Classifier. In ball tracking, we calculate likelihood edge and color feature. The player's points project to pitch by projection transform. Experimental results show the effectiveness of the proposed tracking method.