Wheelchair-user Detection Combined with Parts-based Tracking

In recent years, there has been an increasing demand for auto m tic wheelchair-user detection from a surveillance video to support wheelchair users. However, it is diffi cult to detect them due to occlusions by surrounding pedestrians in a crowded scene. In this paper, we propose a de tection method of wheelchair users robust to such occlusions. Concretely, in case the detector cannot a d e ect wheelchair user, the proposed method estimates his/her location by parts-based tracking based on par ts relationship through time. This makes it possible to detect occluded wheelchair users even though he/she is he avily occluded. As a result of an experiment, the detection of wheelchair users with the proposed method a chieved the highest accuracy in crowded scenes, compared with comparative methods.