Research on Athlete Image Detection and Shadow Removal Algorithms in Tennis Matches

To validate a method proposed in this paper that combines athlete detection with an improved Gaussian mixture model to remove shadows of athletes in tennis matches, the following steps were undertaken. Initially, the circular detection method was applied to detect athletes in the match images. Subsequently, the area of background modeling was reduced based on the identified athlete positions. Addressing the issue of slow movement during tennis serve and volley, which leads to easy “missed detection,” an improved Gaussian mixture model for shadow detection using the eight-neighborhood method was proposed. Finally, the image was converted to the HSV color model, and the difference in feature values of shadow information in this color model was utilized to achieve shadow elimination. Comparative analyses were conducted to evaluate the effects of this method, Gaussian mixture modeling, and background difference modeling on shadow elimination. Additionally, shadow removal experiments were performed under different lighting conditions. The experimental results demonstrate the effectiveness of the proposed method in removing shadows of athletes in tennis matches.