Motion recognition from contact forces information and identification of the human body dynamics

The knowledge of the segment parameters (SP) of the human body is crucial to understand, analyze and simulate human motions. The most recent technologies have enabled to identify the segment parameters, to compute in real-time variations of the Total Center of Mass (TCoM), or the Center of Pressure (CoP). These methods make use of costly and precise equipments, not easily portable for in-house personal use. In this paper we investigate the possibility to recognize motions from the contact force profile for a variety of 26 gymnastic movements; and then to identify the body segment parameters making use of the motion information (joint angles, velocities and accelerations computed from inverse kinematics of passive optical markers 3D coordinates) stored in a database. The method is tested with 3 candidates of different level of expertise in the realization of the prescribed motions.