Basketball Foul Model and Judgment System Proposal

Recently, the practical applications of artificial intelligence (AI) in various fields have increased owing to advancements in AI technologies, such as image recognition and machine learning. Video judgment is often used in ball games and athletics for arbitration, e.g., the Hawk-Eye system in tennis. However, bad decisions are a frequent problem in sports, and the arbitration results tend to differ based on the subjectivity of the human referees; hence, a system that ensures objective judgment is required. In this paper, we propose a solution to judge fouls in basketball automatically. First, the players are designated as research objects; then, the objects in basketball fouls are classified, and recognizable motions are defined for each category. Next, a basketball foul model was designed, and current motion information of each object was extracted from real-life basketball match videos. Finally, a basketball foul judgment system was designed to extract the offending objects and display them in the video in real time.