Intelligent Recognition of Erroneous Movements in Athlete Training Based on Computer Vision Technology

Human motion recognition has great practical value in the field of computer vision, but it is also extremely complex and a hot spot and difficult area of computer vision research. The purpose of this paper is to study the intelligent recognition of erroneous movements in athlete training based on computer vision technology. The principle of YOLO algorithm is analysed, and a window segmentation method based on a combination of sliding window and action window is proposed for action recognition based on a double-layer classifier in terms of mobile data window segmentation, and applied to an intelligent error action recognition system. Finally, it is demonstrated that the method is able to accurately identify action types with a final false action recognition rate of 95%.