Nonlinear State-Space Modeling of Human Motion Using 2-D Marker Observations

A novel method for the estimation of human kinematics, based on state-space modeling, is proposed. The state consists of the positions, orientations, velocities, and accelerations of an articulated model. Estimation is performed using the unscented Kalman filter (UKF) algorithm with a fixed-interval smoother. Impulsive acceleration at floor contact of the foot is estimated by implementing a contact constraint in the UKF evolution model. The constraint inserts an acceleration impulse into the model state. The estimation method was applied to marker-based motion analysis in a motion laboratory. Validation measurements were performed with a rigid test device and with human gait. A triaxial accelerometer was used to evaluate acceleration estimates. Comparison between the proposed method and the extended Kalman smoother showed a clear difference in the quality of estimates during impulsive accelerations. The proposed approach enables estimation of human kinematics during both continuous and transient accelerations. The approach provides a novel way of estimating acceleration at foot initial contact, and thus enables more accurate evaluation of loading from the beginning of the floor contact.