Compressive Sensing Based Indoor Occupancy Positioning Using a Single Thermopile Point Detector With a Coded Binary Mask

Abstract

Compressive sensing based indoor human detection has raised accumulative attention because it could recover sparse signals from fewer samples than that needed by the Nyquist sampling theorem. However, their application is limited due to high computation cost and large deployment effort. To address this, this article proposes the use of a rotating binary mask and a single thermopile point detector to enable compressive sensing for indoor human positioning. Traditionally, a thermopile point detector could only convert the received infrared radiation into electrical signals. By using a periodically rotated coded binary mask, the infrared radiation received by the detector will be sampled compressively, and a compressed signal sequence will be obtained after a few measurements. The mask consists of eight submasks, and seven of them are designed following the pattern of a randomly generated binary matrix. A recovery algorithm is implemented to recover the original spatial distribution of the infrared radiation within the field of view. Experimental results demonstrate the zone-level positioning with high accuracy. The proposed sensor is compact, mobile, cheap, and easy to deploy and, therefore, has great potential for occupancy state monitoring, such as occupant centered thermal comfort control.