Biomechanics Analysis of Sepak Takraw Tekong Serves via Depth Camera Motion Capture System

This study is focussing on the biomechanics analysis of the Sepak Takraw Tekong pre-serve motion. To date, most biomechanics analysis in sports is using standard three-dimensional video camera recording as the motion capture system . In this project, an in-house marker-less three-dimensional (3D) Depth Camera motion capture system is introduced. Kinect for Xbox 360 is chosen as the 3D Depth Camera motion capture sensor. Two sensors are used for duplex motion capture system arrangement since a recent study has proven that duplex system provides better results as compared to single Kinect motion capture system. These sensors have the capability to capture motions at 30 fps and categorized as marker-less skeleton tracking system. A server or known as Tekong in Sepak Takraw sport was chosen as test subject and his motion during performing the pre-serve motion was analysed. The serves motion of Tekong was captured during training session. Total serves attempt captured were 15 attempts. Data collected from the Kinect was then analysed using Matrix Laboratory software. The parameters analysed were foot-to-foot distance and both knee relative angle at pre-serve phase. The mean value and standard deviation for foot-to-foot distance in this study are 0.58 and 0.19 m. For the biomechanics analysis on the sample chosen, the mean value for relative angle for right knee was 148.72° whereas left knee relative angle was 160.89°. Speed of the ball being served also recorded in this study. From the statistical analysis, the result shows that the foot-to-foot distance and relative angle of the knee during pre-serve phase have an effect on the speed of the ball being served. From the outcomes of the study, it is suggested to use Kinect as the motion capture system since it can provide marker-less which is unable to comfort the subject to be examined, and also it can provide data in three dimensional. But, improvement in the specification of Kinect sensor used was highly recommended.