Automatic Video-based Analysis of Athlete Action

Video analysis of athlete action is becoming an important tool for sports training, since it has no intervention to the athlete and there are abundant archived data it can exploit. In this paper we present our work on the automatic analysis of complex individual actions in diving video aiming at providing biometric measurements and visual tools for coaching assistant and performance improving. The main body joint angles of the athlete are automatic obtained by 2D articulated human body model fitting and shape analysis techniques. Two visual analyzing tools: motion panorama and overlay composition, which are extremely suitable for individual sports game training are presented. The encouraging experimental results show the effectiveness of the proposed system.