High SBP compressive holography with an efficient block-wise model

Abstract

High space-bandwidth product (SBP) imaging platform is an indispensable technique in diverse research fields, particularly in medical imaging and diagnosis. Because of the scale-dependent geometric aberrations of optical elements and the limited number of pixels of image sensor, high resolution and large field of view are hard to be realized simultaneously in traditional imaging systems. Compared with off-axis holography, in-line lens-less holography could maintain high SBP without an imaging lens. Compressive holography (CH), which combines compressive sensing and in-line lens-less holography, is considered as a promising solution for high SBP three-dimensional imaging. We developed a high SBP three-dimensional imaging algorithm using CH based on total-variation sparsity constraint. An efficient block-wise CH algorithm is proposed to reduce the reconstruction time. The block-wise model could locate accurate reconstruction searching spaces, resulting in high convergence speed and high image contrast.