A multi camera system for soccer player performance evaluation

This paper presents a multi-view approach to performance evaluation of soccer players by the analysis of the posture evolution. Some body-appearance features have been extracted and the most significant ones have been used to model the activity of the players involved in play. Continuous hidden Markov models have been used to model the temporal evolution of the body features in a multiple view decision making approach. Tests were carried out on different sequences of player activities extracted from matches played during the Italian ldquoSerie Ardquo Championship.