Estimation of the number of humans and their movement paths in a room using binary infrared sensors

It is very important to be able to determine humans' locations and their movement in order to a ubiquitous service environment. In this paper, we propose an algorithm that simultaneously estimates the number of humans and the movement path for each human in a room, using only the binary sensing data obtained from infrared sensors attached to the ceiling. Compared to other camera-based systems, the estimation results of our algorithm can be used directly in application systems without violating anyone's privacy. After describing the proposed algorithm in detail, we then describe our evaluation results.