An application for spatiotemporal analysis of human gait has been developed in MATLAB. It allows for recording sequences of depth images using a Microsoft Kinect v2 sensor and estimating selected spatiotemporal parameters characterizing the human gait recorded on the footage.

An algorithm for processing data representative of positions of feet has been implemented. The first step of the algorithm is to acquire the coordinates of the feet position during the transition recorded on the video and to smooth the course of these coordinates. The next step is to differentiate the foot trajectories in order to obtain the instantaneous foot velocity values. The instantaneous velocity values are then compared with a threshold to determine whether the feet are in motion or at rest at given moments. After this operation, the time moments corresponding to the moments of foot contact with the ground and the moments of foot detachment from the ground are determined. The last step is to calculate the values of the spatiotemporal parameters based on the determined time moments and the coordinates of the feet position in these time moments.

The effectiveness of the implemented algorithm for the estimation of selected spatiotemporal gait parameters was verified by comparing its results with the results obtained using a Zebris pressure distribution measurement platform.