Backpack: Detection of people carrying objects using silhouettes

We describe a video-rate surveillance algorithm for determining whether peo ple are carrying objects or moving unencumbered from a stationary camera. T he contribution of the paper is the shape analysis algorithm that both dete rmines whether a person is carrying an object and segments the object from the person so that it can be tracked, e.g., during an exchange of objects b etween two people. As the object is segmented, an appearance model of the o bject is constructed. The method combines periodic motion estimation with s tatic symmetry analysis of the silhouettes of a person in each frame of the sequence. Experimental results demonstrate robustness and real-time perfor mance of the proposed algorithm. (C) 2001 Academic Press.