3D fall detection for single camera surveillance systems on the street

Suneung Kim, Myeongseob Ko, Kyungchai Lee, Mingi Kim, Kwangtaek Kim

Abstract: As the number of older persons has increased substantially in recent years in most countries, researchers and scientists pay more attention to develop automatic fall detection technologies. In this paper, we introduce a new method that provides accurate fall detection under both indoor and outdoor environments by using depth images generated from a single image sequence with a machine learning algorithm. For the fall detection, we developed an Extended Kalman Filter based 3D human tracking that utilizes both 2D and 3D information of a dynamic scene. Due to the benefit of depth information, our method detects and tracks a moving human accurately without having background subtraction. Our solution is a promising technology for surveillance camera systems on the street.