Practical application of interactive AI technology based on visual analysis in professional system of physical...

Currently, physical education teaching in universities tends to adopt the traditional model of one teacher for multiple students, which has high requirements for teachers and is difficult to consider students' strengths. On this basis, an interactive system has been established, including three modules: behavior information, user data collection, and behavior evaluation. Taking the 400 m running physical education teaching as an example, Kinectv2 was used to collect students' movements and contours while running, and ORB (Oriented FAST and Rotated BRIEF) feature extraction algorithm was used to extract students' movement features. After importing the data into the interactive system, students and teachers could view it in the system and provide guidance based on the students' actions. This article took 10 students as examples to test their performance changes in the 400 m running before and after systematic training. The results showed that the evaluation score after receiving systematic instruction increased by 6-7 s compared to the score without receiving instruction, with a significant change. This indicated that the interactive AI (artificial intelligence) system constructed in this article can play a significant role in sports teaching of 400 m running.