Snapshot multiplexed imaging based on compressive sensing

Abstract

Multiplexed imaging methods have been proposed to extend the field of view (FoV) of the imaging devices. However, the nature of multiple exposures hinders its application in time-crucial scenarios. In this paper, we design a snapshot multiplexed imaging system for wide FoV imaging. In the system, the scene is first spatially encoded by a mask, and then the coded scene is optically divided into multiple sub-regions which are finally superimposed and measured on a sensor array. We model the demultiplexing as a compressive sensing (CS) reconstruction problem and introduce two methods, one is based on Total Variation (TV) constraint and the other is based on sparsity constraint, to reconstruct the scene. Simulation results demonstrate the effectiveness of the proposed system.