Sleep recognition algorithm using depth sensors

Abstract

With the advance of natural user interface technology, studies on detecting human movement and gestures have been conducted. Among these studies, sleep recognition is significant for energy saving as human's special behavior. This paper proposes a novel approach using a depth sensor and camera for increasing sleep recognition accuracy. Several potential errors of sensing gestures when sleeping are solved by the proposed algorithm. The experiment shows that the proposed approach achieves high sleep recognition accuracy. This approach is practically meaningful for application of energy saving and security systems. © 2012 IEEE.