Feature extraction of basketball player's foul action using machine vision

The rising demand for basketball parallels the advancement of technology and the global economy, enhancing people's lifestyles. Modern basketball emphasises height, speed, accuracy, and strength, with increased physical contact. This intensifies gameplay, leading to inevitable mistakes. Foul regulations add unpredictability and challenge for players, coaches, and referees, influencing game outcomes. This study employs machine vision to analyse basketball foul movements. Experimental findings reveal patterns: the Chinese team committed 16 dribble defence fouls and 41 shooting defence fouls, while opponents committed 33 and 32, respectively. For ball defence, the Chinese team fouled five times in both preventing and defending, and four in pass prevention, compared to opponents' seven, 14, and five fouls, respectively. These results underscore the potential of machine vision in detailed detection of player fouls, offering insights for basketball development and gameplay refinement.