Heel strike detection based on human walking movement for surveillance analysis

Heel strike detection is an important cue for human gait recognition and detection in visual surveillance since the heel strike position can be used to derive the gait periodicity, stride and step length. We propose a novel method for heel strike detection using a gait trajectory model, which is robust to occlusion, camera view, and low resolution. When a person walks, the movement of the head is conspicuous and sinusoidal. The highest point of the trajectory of the head occurs when the feet cross (stance) and the lowest point is when the gait stride is the largest (heel strike). Our gait trajectory model is constructed from trajectory data using non-linear optimisation. Then, the key frames in which the heel strikes take place are calculated. A Region Of Interest (ROI) is extracted using the silhouette image of the key frame as a filter. For candidate detection, Gradient Descent is applied to detect maxima which are considered to be the time of the heel strikes. For candidate verification, two filtering methods are used to reconstruct the 3D position of a heel strike using the given camera projection matrix. The contribution of this research is the first use of the gait trajectory in the heel strike position estimation process and we contend that it is a new approach for basic analysis in surveillance imagery.

► We detect heel strikes using a gait trajectory model under visual surveillance.
► This method is robust to occlusion, camera view, and low resolution.
► When a person walks, the movement of the upper body is conspicuous and sinusoidal.
► The model is constructed from trajectory data using non-linear optimisation.
► It is the first use of the gait trajectory in the heel strike position estimation.