Multi-target segmentation in a multi-camera surveillance environment

This paper presents a multi-target segmentation system in a multi-camera surveillance environment, which can be used in smart video surveillance applications. The system is composed of three subsystems, i.e., the foreground and background segmentation subsystem (FBSS), the human body detection subsystem (HBDS), and the human feature matching subsystem (HFMS). By means of the integration of these subsystems, moving body detection in the multi-sensor images of the same scene using video cameras is performed. Experimental results show that the proposed system provides acceptable moving body detection rate.