Compressed infrared bearing sensor for human localization: Design and implementation

Abstract

This paper presents the design and implementation of a compressed infrared bearing sensor for human localization. The developed bearing sensor consists of multiple planar co-located pyroelectric infrared (PIR) detectors with the same planar sensing coverage. The observed space of the bearing sensor is defined by partitioning the sensing area into discrete fan-shaped sampling cells. The spatial mapping from the observed space to the measurement space is built by applying the visibility masks to the Fresnel lens arrays associated with the PIR detectors. We explore the use of a compressive sensing paradigm in designing visibility masks with desirable sensing efficiency. The proposed bearing sensor is experimentally validated in the context of the cross-bearing localization of a human target with a mobile robot.