This work combines the kinematics estimate of human standing with a hybrid identification algorithm to identify a set of ankle dynamics mechanical parameters. We used the hold and release (H&R) experimental paradigm to model a set of recoverable falls on a population of unimpaired adults. Body kinematics was acquired with a microsoft kinect (mk) version 2 after benchmarking its position accuracy to a camera-based vision system (CVS). The system identification algorithm, combining an extended Kalman filter (EKF) and a genetic algorithm (GA), allowed to identify the effect of tendon and muscle stiffness at the ankle joint, separately. This work highlights that, when associated to soft-computing techniques, affordable tracking devices developed for the gaming industry can be used for the reliable assessment of neuromechanical parameters in clinical settings.