Aiming at the fundamentals and importance of sports action technology, combined with the existing development of sports action image recognition and analysis technology, this paper introduces the corner mechanism of quantum revolving door into the quantum genetic algorithm, and proposes the gradient-based adaptive quantum genetic algorithm. Using bilinear interpolation method to geometrically transform the sports action images, combining the normalized interrelationship measure with the improved quantum genetic algorithm to realize the sports action image matching process. To statistically compare the performance of the improved quantum genetic algorithm with other algorithms for sports action image segmentation. Analyze the total mean scores of teaching ability of sports trainee teachers when using quantum genetic algorithm to adjust sports movement techniques. In the function test, the improved quantum genetic algorithm adopts the quantum gate adaptive rotation angle strategy, which can avoid algorithm divergence or premature convergence, improve the accuracy and speed of algorithm optimization, and make the algorithm more robust. The total mean scores of teaching ability of physical education trainee teachers in the initial stage and the end stage of the internship were 3.49 and 4.45 respectively. The difference of the total mean scores of teaching ability proved the feasibility of the image matching technique of quantum genetic algorithm for the improvement of movement skills and the enhancement of teaching professionalism of physical education teachers.
