Computational photography and compressive holography

Abstract

As lasers, photosensors, and computational imaging techniques improve, holography becomes an increasingly attractive approach for imaging applications largely reserved for photography. For the same illumination energy, we show that holography and photography have nearly identical noise performance. Because the coherent field is two dimensional outside of a source, there is ambiguity in inferring the three-dimensional structure of a source from the coherent field. Compressive holography overcomes this limitation by imposing sparsity constraints on the three-dimensional scatterer, which greatly reduces the number of possibilities allowing reliable inference of structure. We demonstrate the use of compressive holography to infer the three-dimensional structure of a scene comprising two toys.