Research on teaching basketball free throw action process based on motion vector field transformation

Abstract

Abstract Free throw shooting is an important training program for athletes, as well as the most effective way to score in the game. The study uses the 3DRS model to figure out the initial motion, fixes the vectors by figuring out the weighted bidirectional motion and smoothing the motion vectors, adds adaptive motion compensation, and suggests a way to look at training videos for free throw shooting that is based on changing the motion vector field. We test different video sequences using this paper’s algorithm and other algorithms, and compare the effects of free throw prediction before and after the motion vector field transform algorithm. We also select basketball players to conduct control experiments to investigate the effectiveness of this paper’s motion vector field transform algorithm. Our research shows that the average PSNR and SSIM values for different sequences are higher than those of other algorithms by 6.16% to 7.00% and 4.50% to 5.00%, respectively. Using the motion vector field transform also makes free throw prediction 4.3% more accurate overall. Another thing is that the athletes who were taught using the motion vector field transformation algorithm got 19.23% and 16.07% better at hitting free throws and other technical measures. This shows that the algorithm in this paper can be used in real life to teach basketball free throw shooting moves.