Video-Based People Fall Detection via Homography Mapping of Foreground Polygons from Overlapping Cameras

In this paper, we investigate a video-based method of detecting fall incidents from overlapping cameras. Our aim here is to propose a novel method, without any wearable device, to detect falls on the floor with a multiple cameras system by using homographic projection on a ground plane (or on reference camera view plane). Two relatively orthogonal views are utilized, in turn, simplifying the estimation of the surface of the person which is in contact with the ground according of the foreground information of each camera. This information is computed in order to differentiate lying on floor posture which can be considered as fall to other position. The performance of our method is tested on a public multi-view fall dataset. The results show the accuracy of our proposed algorithm.