Design and Implementation Strategy of Informative Training System for Tennis Physical Education

Tennis teaching generally suffers from the problems of insufficient professional teachers and backward teaching methods, which results in students not being able to get the correct guidance and tennis courses not being able to achieve the desired teaching effect. Therefore, this paper proposes to apply the methods of posture estimation, movement recognition and movement evaluation in the teaching and training of tennis, and design the tennis sports informatization training system, so as to effectively improve the efficiency and quality of tennis teaching. The system employs the OpenPose-bm algorithm to estimate pose, extracts the skeletal pose point characteristics of the target character, and acquires the video skeletal pose point sequence of the target character. The skeleton action recognition network (AA-GCN) method is utilized for recognition, to classify and recognize the actions of the tennis player based on the coordinate information of the key points of the human skeleton. Finally, the similarity algorithm is used to provide recognition scoring guidance for tennis sports actions. The experiment shows that the method of this paper has achieved an accuracy rate of 0.8889 in the task of recognizing the six basic skills of tennis. And in the 15-week practical teaching application, it significantly improved the assessment scores of the six basic tennis skills of the students in the experimental group. In this way, it can create a personalized guidance plan for athletes’ daily training based on the information-based training system.