Viscoelastic properties of the ankle during quiet standing via raster images and EKF

The present study deals with the reconstruction of the continuous-time state space parameters proper of human quiet standing. The reconstruction utilized a hybrid non-linear extended Kalman filter to combine a biomechanical model with the discrete-time position measurements provided by two web-cameras via a General Linear Camera model. After camera calibration and validating the filter in simulation, we performed the estimation on a group of volunteers whose quiet standing was perturbed by means of a hold and release paradigm. The filter allowed estimating stiffness and damping of the ankle during quiet standing as well as the kinematic variables of the subjects´ center of mass.