Spatio-Temporal motion analysis based suspicious behavior detection

Suspicious behavior detection has been developed in an unsupervised way based on spatio-temporal motion analysis in pedestrian virtual space. The pedestrian virtual plane has been constructed which consists of both the ground plane and the pedestrian head one. The suspicious behavior has been discriminated in a circular variance of pedestrian trajectories around the 3D virtual region instead of traditional 2D protected one. The protected region can be assigned as different shapes and sizes. Experiments have been shown that the proposed approach is efficient for distinguishing the anomaly in a protected region without any hypothesis for the scenario contents in advance.