Validation of ball spin estimates in tennis from multi-camera tracking data

Area of Science:

• Sports Science
• Biomechanics
• Motion Analysis

Background:

• Accurate measurement of ball spin is crucial in many sports for performance analysis and strategy.
• Multi-camera tracking systems offer a non-invasive method for capturing ball trajectory data.
• Existing technologies like Hawk-Eye have limitations in precisely estimating spin parameters.

Purpose of the Study:

• To evaluate the accuracy of a theoretical ball trajectory model compared to Hawk-Eye for estimating ball spin rate and spin axis direction.
• To determine the effectiveness of multi-camera tracking technology in capturing spin data.

Main Methods:

• Comparison of a theoretical ball trajectory model and Hawk-Eye's method against high-speed video analysis.
• Assessment of accuracy using metrics such as mean bias, standard deviation, limits of agreement, and Root Mean Square Error (RMSE).

Main Results:

• The theoretical ball trajectory model demonstrated superior accuracy in estimating ball spin rate and spin axis direction compared to Hawk-Eye.
• The trajectory model showed a lower mean bias (2.92 ± 222.76) and standard deviation (-100.01 ± 542.44) than Hawk-Eye.
• Higher accuracy was achieved for spin rates below 4500 RPM; however, neither method consistently estimated spin rates above 4500 RPM accurately.

Conclusions:

• A theoretical ball trajectory model applied to multi-camera tracking data provides a practical and accurate method for estimating ball spin.
• This approach facilitates large-scale, non-invasive collection of spin rate data during sporting events.
• Further research may be needed to improve accuracy for very high spin rates.