Design of Assistant System for Basketball Shooting Training of Teenagers Based on Deep Learning

At present, there are some phenomena in the development of campus basketball, such as insufficient training venues, low coaching level of coaches, imperfect competition system and unbalanced regional development, which will inevitably affect the development of campus basketball characteristic sports. Shooting is the main offensive technique of basketball and the only means to score points. Therefore, in order to meet the requirements of automatic action evaluation in basketball, this paper develops a teenagers’ basketball shooting training assistant system based on deep learning, and uses YOLOv4 to classify the influence of basketball frame by frame. In order to analyze the shot information and flight trajectory of basketball, this paper realizes basketball detection by improving Hough transform. The research results show that compared with the traditional training methods, the fine movements obtained by the teenagers’ basketball shooting training assistant system in this paper have certain advantages and can bring better teaching effect.