Suspicious behavior detection of people by monitoring camera
Wassima Aitfares, Abdellatif Kobbane, Abdelaziz Kriouile · 2016
The analytic video is a very challenging area of research in computer vision. Ensure a high level of security in a public space monitored by a surveillance camera is a difficult task in recent years. Understanding people behaviors in real time allows the surveillance systems to analyze unusual events through the video frames. In this paper, we propose a new approach for detecting suspicious behavior of moving people. We are not interested in a simple motion detection of a moving object, but we analyze the trajectory of this latter; relying on the object motion vector. Once a suspicious behavior suddenly occurs in this trajectory, we segment and track this object during its motion within the camera's field of view. Experiments with real-world images validate the efficiency of the proposed approach.