Vision-based Upper Extremity Kinematic Analysis of Badminton Smash Hit

Abstract

Badminton is one of the fastest sports in the world. Athletes’ body rotations introduce a lot more complexities for studying badminton. The goal of this work is to develop a series of analysis including designing MATLAB programs and acquire kinematics data of badminton sport to help badminton players improve their skills. To validate these analyses are consistent with what badminton coaches teach athletes, interviews are conducted with coaches at the collegiate level to set commonly agreed “coaching points”. A full body marker setup is applied to this study while the markers are hand-made to reduce the weight and improve accuracy. 6 badminton players in different skill levels are brought in for data collecting. Subjects are asked to perform a smash motion as a return of a high and long badminton serving. Smashing data was captured by an eight-camera Vicon Motion Capture System (Nexus 1.8.5), and one video camera. Analyses were programmed by MATLAB to examine coaching points: speed and acceleration of shuttlecock and racket, contact point on racket net surface, coefficient of momentum transfer, coefficient of restitution, the angle between racket surface direction, racket moving direction, and angular velocity contributions from body parts. The results indicate the experiments and analysis are successful. Most of the ”coaching points” are proved by comparing kinematic quantities from players. These kinematic quantities can be applied as indicators to show the skill level of a player. With further development of this research, the coaching system can provide a series of dynamic analysis and lively visualized feedback to the athlete.