Trajectory reconstruction system of moving target applied in volleyball match

Abstract

For the tactical analysis in sports video, the trajectory of ball and players is very valuable data. Therefore, this paper puts forward a system of reconstructing the trajectory of ball and players from volleyball videos. The main tasks include player detection and tracking, trajectory repair and Volleyball scene 2D mapping. According to the fixed competition area, the Yolov5 algorithm is used to train the detector, and vol-DeepSORT is combined to realize the real-time detection and tracking of the players, which can reduce the problems of missed detection and false detection caused by long-term occlusion and lens running. Finally, the camera calibration is used to reconstruct and display the trajectory information of the ball and players, which is helpful for professionals to know effectively and comprehensively the tactical application in volleyball matches.