Validation of a real-time video analysis system for soccer

The aim of the current study was to determine the validity of an automated multiple-camera tracking system (Venatrack™), for the measurement of speed during soccer. 18 recreationally active males (mean age 27±8.6 years) performed a series of runs (total 391), representative of the movements made by soccer players, at given speeds. The speeds recorded by the automatic tracking system were compared statistically with speed measurements made using timing gates. For all the runs combined the mean speed recorded by the automated system was 15.4±5.5 km·h - 1 compared with the recorded mean speed of 15.2±5.4 km·h - 1 and the mean difference and 95% limits of agreement were - 0.25±0.64 km·h - 1. Pearson correlations (r) among timing gate speed and automated tracking speed were ≥ 0.99 (P<0.001), except the 20 m sprint, with 90° turn (r > 0.7). For the zig-zag shuttle the mean speed recorded by the automated system was 21.2±3.5 km·h - 1 compared with the recorded mean speed of 20.8±3.4 km·h - 1. The results demonstrate good validity over a range of soccer specific movements and speeds, up to and including sprinting. The results of this study suggest that the automated system (Venatrack™) is a valid real-time motion analysis system for tracking player movements during soccer.