Validation of spatiotemporal tennis ball trajectory tracking technology

The introduction of spatiotemporal data capture technology in professional tennis has extended the possibilities of analysing tennis match play. The most direct application of this technology has been electronic line-calling, which removes human adjudication in determining whether a ball lands in play. The suitability of this technology in professional tournaments has therefore focussed on evaluating its accuracy in correctly adjudicating line-calling, with less scrutiny applied to the overall ball flight. However, an accurate prediction of ball trajectory can significantly extend the potential applications of collected spatiotemporal data. This paper presents experimental results and associated methodology for comparing and assessing reconstructed ball trajectories from a commercial electronic line-calling system provider, used in Grand Slam tennis, with those from the Vicon 3D motion analysis system. A tennis court used for professional match play was configured to use both technologies simultaneously, allowing for the recording of player and ball motions of two subjects engaged in tennis rallies. Generated ball motion time histories from both technologies were then spatially and temporally synchronised to compare the average separation in the derived ball trajectories. The average spatial separation of the ball trajectory between the commercial technology and Vicon was 36.8 mm, with the separation being higher at mid-trajectory locations than near landing locations.