Real Time Prediction of Rebound Trajectory of Rotating Table Tennis based on Binocular Vision

摘要

With the deepening of the research on table tennis, intelligent table tennis trajectory prediction and positioning has become a hot issue in the research. The traditional rotating table tennis trajectory prediction method has poor ability to divide the background of table tennis, resulting in poor accuracy of rebound trajectory prediction. Therefore, a real-time prediction method based on binocular vision is designed. The binocular vision model is constructed to obtain the three-dimensional model of table tennis. According to the color space theory, the motion model and background subtraction method are combined to obtain the moving area of the rotating table tennis target. The rebound trajectory of the rotating table tennis ball is predicted in the target active area. So far, the design of real-time prediction of the rebound trajectory of rotating table tennis ball based on binocular vision is completed. The experimental results show that the prediction accuracy of the proposed method is high, which provides a theoretical reference for future research.