How accurate can the automated tennis line-judging system called be? Surely for the level of accuracy it seems to offer, it would need far more cameras than appear to be present at major tennis tournaments. Yet everybody happily accepts its rulings. How does it work?
鈥� Briton Paul Hawkins created and named Hawk-Eye, a system which combined the expertise he gained for his PhD in artificial intelligence with his passion for sport, particularly cricket.
In cricket, a batsman can be given out 鈥渓eg before wicket鈥�. This ruling is applied when the umpire believes the ball would have gone on to hit the stumps had the batsman鈥檚 leg not been in the way. In this situation Hawk-Eye can be used to predict the ball鈥檚 trajectory and is arguably more reliable than an umpire.
Advertisement
Despite being invented with cricket in mind, it was tennis that was receptive to the technology much earlier, perhaps thanks to TV replays showing that umpiring mistakes contributed to the defeat of Serena Williams by Jennifer Capriati in the 2004 US Open quarter-finals. Hawk-Eye provides an instant replay of crucial shots and has also proved an excellent tool for analysing the strategy and performance of players.
For tennis, it relies upon a maximum of six cameras to provide data for sophisticated triangulation. The position of the ball is tracked via a succession of stills from each camera. Within a virtual recreation of the tennis court, a ray can be drawn from each camera through the centre of the ball. The intersection of these rays provides the position of the ball in three dimensions and, with the passage of each frame, its velocity. This can be used to calculate the contact area of the ball with the court, taking into account the distortion of the ball after it is hit. Hawk-Eye also captures any skidding of the ball on the court, which can deceive the eye into believing a ball is out.
Mike Follows, Willenhall, West Midlands, UK