Detect the Body Behavior Characteristics in Sports Training Based on Machine Vision Images
Abstract
To verify its practicability, this paper carried out experiments on it from the three dimensions of detection quantity, detection accuracy, and detection time and compared it with traditional detection methods. The experimental results showed that the average recognition rate of machine vision images in 12 kinds of body behavior image sequences was 83.17%. From this data, it can be seen that the method of behavior feature detection under motion training based on machine vision image is more feasible and can effectively enhance the accuracy of feature detection.
Fields of Study & Subjects
Human Pose and Action Recognition
Sports Dynamics and Biomechanics
Physical Education and Training Studies
Publication details
- Journal / source
- International Journal of Cognitive Informatics and Natural Intelligence
- Volume
- 20
- Issue
- 1
- Pages
- 1–18
- Publication date
- 12 January 2026
- Language
- NG