SYSTEMATIC OF MUSCULOSKELETAL INJURY TYPES AND REHABILITATION TRAINING STRATEGIES FOR ATHLETES BASED ON SPORTS BIOMECHANICS

Abstract

There is an urgent need for sports injury prevention and rehabilitation training, especially in accurately assessing athletes' musculoskeletal parameters and injury types. Based on sports biomechanics, the study investigates the types of musculoskeletal injuries and rehabilitation training strategies for athletes. The study combines the semi-global block matching algorithm, egg-shaped surface model, stereo vision technique, and Hill muscle model to extract key data and injury types of athletes' musculoskeletal. The results of the study indicated that the semi-global block matching algorithm had the lowest mean absolute error in parallax estimation at 0.10. The mean squared error was also the lowest at 0.13, showing superior stability and accuracy. The muscle strength prediction based on Hill model had an error of 0 with actual measured value and a relative error of 0%, which was much better than other models. The SGBM algorithm performed well in terms of both pixel level accuracy and pixel point registration rate. In particular, it quickly reached high efficiency at the beginning of the iteration, showing its superiority in the stereo matching task. The proposed technique can provide an accurate and reliable quantitative analysis tool for the assessment of musculoskeletal injuries and rehabilitation training of athletes, which has important clinical and sports training applications.