Recognition Model Of Sports Athletes’ Wrong Actions Based On Computer Vision
Download as PDF
Computer vision analysis and image processing technology are widely used in the structural analysis of human body. The purpose of this study is to establish a movement recognition model based on computer vision system. In this paper, we establish a three-dimensional coordinate system and formally describe the point inference rules of the flying leg. In this study, by establishing the error function and the deduction model of the flying feet based on the computer vision system, the wrong actions of the head, shoulders, wrists and feet in the skull can be identified clearly, and have a high degree of coincidence with the correct actions. The experimental results show that the accuracy of the algorithm in the 30th experiment is as high as 95%, 15% higher than the traditional algorithm.
Computer Vision; Sports Athletes; Movement Recognition; Model Construction