Estimating Trajectory of Inhabitants with Sparsely Aligned Infrared Sensors

Abstract

We propose in this paper a method for estimating trajectories of the inhabitants in a home environment, which exploits the synergy between location and movement to provide the information necessary for intelligent home appliance control. Our goal is to carry out accurate movement estimation for multiple people in a home environment. We propose an approach that uses information gathered using only passive infrared sensors commonly found in lighting control systems. No special devices or video cameras are needed. Moreover, it is not necessary to carry out data collection for training. We evaluated our approach by conducting experiments in a real home fitted with sensors and we confirmed that trajectories were almost completely detected for four inhabitants who moved upon scenarios.