Player Detection Algorithm Based on Color Segmentation and Improved CamShift Algorithm

摘要: Aimed at the most proposed methods that detecting and tracking moving object cannot receive a good segmentation of player in dynamic scenes, chapter put forward an algorithm, based on color CamShift to detect tennis player. Firstly, apply supervised clustering binary tree automatically target’s area. Secondly, take detected area as initial window, track target by using improved algorithm. On basis accurate tracking, raises new extraction method player, taking use characteristic mostly court’s is simple. The thought extract feature, it combines algorithm with difference frame overcome tracking-loss problem. Experimental results show our effective scenes.