Simulation System of College Physical Education Based on Support Vector Machine Algorithm

Teaching simulation system can not only assist traditional teaching and training methods, but also stimulate athletes' interest in learning and training, and improve their ability to recognize the main role. Therefore, physical education needs the help of high technology to meet its own needs. Based on this, it is an urgent and necessary work to build a physical education simulation system. In this paper, a simulation system of college physical education based on SVM(support vector machine) algorithm is designed. This physical education simulation system can finally realize the effects of real-time control, all-angle switching and multi-angle display of teaching contents in physical education through kinect equipment. And it can be displayed from all directions and angles according to the three-dimensional model. In this paper, a new idea based on hierarchical classification is used. According to the motion description, the motion is separated in advance, and then the SVM-based method is used for training. It is transmitted frame by frame in the way of retraining, which provides more and more complete historical information for reference for moving target tracking. The experimental results show that the algorithm has strong robustness in some video image sequences with drastic target changes. The recognition rate of human body movements in sitting position is higher, and the physical education simulation system designed in this paper can improve the quality of physical education.