Spatio-temporal analysis of kinematic signals in classical ballet

Motions of markers arranged on a dancer's body can be approximated by the sum of a minimal set of linear trajectories with given accuracy. The composition of approximating linear trajectories features the movement traits and discloses the level of movement expertise in the dancers. We suggest the computationally simple methods for the analysis of trajectories and body shape changes attested directly from the motion capture data. We have tested our approach for 6 figures from the classical ballet repertoire performed by 24 dancers varying in expertise. The methods allow to estimate the level of movement expertise, to draw the detailed structure of movements, and to classify movements into a given repertoire automatically.