Computer vision approach to automatic linesman

摘要

To date, there is no implementation of an automatic linesman system (ALS) in any official badminton game. An ALS is important in assisting the game umpire/referee to make a better judgment in a dispute. Just like the Hawkeye system in tennis, the ALS is anticipated to play more important role in big games that require high standard refereeing quality like the Olympic, Badminton World Championships (BWF) etc. This paper proposed an algorithm to model the shuttlecock position so as to track and determine its landing point either inside or outside-the boundaries of the court line. It involves background subtraction method in hue space, analysis of the shuttlecock trajectory and a decision making process to determine the shuttlecock landing point. Performance is evaluated based on three metrics namely, accuracy, sensitivity and specificity. Initial results showed that the proposed algorithm works well with low error values. However, further works is required to refine the proposed ALS before it can be fully implemented in an official game.