The Application on Video Analysis in Computer Aided Evaluation for Football Skills Training

摘要

Computer Aided Evaluation has been widely used in various application including physical training and athletic sports. Instead of subjective assessments of a coach the video records and analysis are proposed for estimation in football skill training. The proposed approach involves the real time image acquisition and processing on the motion tracking for a football player. A fisheye lens with CCD camera, which provide a extremely wide range of the scene, are utilized for record of the player's moving action. The Background Differencing Algorithm, which is a combination of the common Background Differencing and the coterminous frames differencing, is employed for the targets extraction and recognition. Then we can test the ordinary players' major actions and find out the best matching with the standard motion template of the professional player. A quantitive analysis will be applied by means of a binary chart generation and calculation with ARTooIKit Algorithm. The comparation will creat and improve the training instruction. Finally, we can get a solution on the skill training improvements.