Exploiting Temporal Statistics for Events Analysis and Understanding

In this paper, we propose a technique for detecting possible anomalous events in an area monitored by a video surveillance system. In particular, here we focus on the time spent by an object to carry out simple events. Mixtures of Gaussians are maintained for each event to have a statistical representation of the times commonly required to perform certain activities. Such statistics are then exploited both for the analysis of the simple activities and for discovering anomalous situations (i.e. complex events). In these cases the system requires the attention of the human operator. Experiments have been performed on a multi-camera system for parking lot security.