An Analysis Model of Sports Human Body Based on Computer Vision Tracking Technology

Abstract

This paper proposes the analysis model of sports human body based on computer vision tracking technology. Visual target tracking is an important research field of the computer vision, motion trajectory and it can provide not only the goal, to provide the initial data movement analysis, scene understanding, behavior or the event detection in intelligent surveillance, human-computer interaction, robot visual navigation and motion recognition based on field has a broad application prospect. For this reason, it is possible to consider the use of a large number of unlabeled samples to assist the training classifier to improve its performance. This type of machine learning method using both labeled and that unlabeled samples is called the semi-supervised learning. This paper proposes the novel idea of the related research topics to propose the new perspective of the model that will be later give us the novel idea of making it efficient for further development of sport science.