Model development for the application of sports data visualization techniques in sports science research

With the continuous development of visualization technology and machine learning algorithms, sports data visualization is more and more widely used in sports training and management. In this paper, the visualization interface based on MFC and the drawing virtual environment of OpenGL is used to display the human motion data visualization function. Meanwhile, an algorithm based on a normalized Euclidean distance matrix is proposed in the process of mapping the sports data to virtual character models with different morphologies in order to maintain the specific relationship between body parts and between the body and the environment space. Applying motion data visualization to the individual performance of sportspersons, the similarity matching algorithm for motion data streams is investigated, and a DTW algorithm based on early stopping with lower complexity and less computation is proposed. The application of sports data visualization technology involves analyzing human movement patterns and proposing a gait prediction method that improves prediction accuracy by 84.8% compared to the prediction method proposed by Alexander. In conducting a comparative study of athletes' movement flow, the scoring results of this paper's sports data visualization technology have a P-value of 0.864 compared with the scoring results of the coach. Cohen's d-value of the magnitude of the difference is 0.087, which is not a significant difference and is able to satisfy the requirements of sports training methods. © 2024 Liang Tan et al., published by Sciendo.