Abstract: This article uses human pose estimation algorithms and Wed technology to develop an intelligent sports action recognition system, summarizes its application in university physical education courses, extracurricular physical exercise, training, and competitions, and innovates the“learning practice competition evaluation”system in university physical education; Using experimental research methods, verify the intervention effects of this system on student evaluation and physical health, improve students'physical health level with new concepts and methods, promote the integration of learning, practice, competition, and evaluation in university sports, and deepen the integration of university sports education and physical education. The results showed that the intelligent sports learning and training competition evaluation system of the university had a positive intervention effect on the evaluation and physical test scores of college students, significantly improving their evaluation scores. Longitudinal analysis of the experimental group test scores showed that the pass rate of students'physical tests was increasing year by year. In the future, further exploration can be conducted to broaden the intervention dimensions and application scope of this system, optimize the application environment of artificial intelligence technology in the field of school sports, and expand the research results to the level of social sports and competitive sports.

Key words: “Artificial intelligence +”, Intelligent sports action recognition system, College sports, Integration of “learning,training,competition,and evaluation”, Student evaluation of teaching, Physical health