Image Processing Techniques for Evaluating and Modeling Athletes' Cognitive Abilities and Sports Performance

Abstract

To delve into the relationship between athletes’ cognitive abilities and sports performance, this article adopted image processing analysis and modeling techniques. By analyzing athletes’ behavior during competitions, it was found that athletes’ cognitive abilities had a significant impact on their sports performance. A professional camera equipment with high resolution, high frame rate, and low noise was chosen to capture moving images from multiple positions and angles. Pre-processing was performed on the collected images. Wavelet analysis method was used for denoising. Residual encoding and decoding network were used for image correction, and grayscale co-occurrence matrix method was used for image feature extraction. In terms of sports accuracy, the distribution of sports accuracy among the 10 athletes in the high-level sports cognitive level group was between 98.02% and 99.99%: in terms of overall reaction errors, the higher the overall reaction error index, the better the exercise performance. The comprehensive response error index of the high-level sports cognition group during 10 basketball training sessions ranged from 3.29 to 3.82, with each index higher than the other two groups; in terms of operation thinking time, the shooting positioning operation time of the high-level sports cognition group was particularly short, and the low-level sports cognition group was on the contrary. The experimental results indicated a significant relationship between cognitive ability and motor performance. High-level athletes performed better in cognitive abilities and their athletic performance was also better.