Compressive imaging for superresolution from a single exposure

Abstract

Superresolution is typically achieved by taking multiple slightly different exposures. Here we present a superresolution technique that requires only a single exposure. Inspired by the compressive sensing approach that takes advantage of the information redundancy in typical images we show that true superresolution is feasible from a single exposure. The single shot superresolution system is based on the well known double random phase encoding setup which was broadly investigated for its encryption purposes. It is shown that combining double random phase encoding imaging with a compressive sensing based reconstruction technique, can super-resolve both diffraction and geometrical limited resolution imaging.