Design of binary patterns for speckle reduction in holographic display with compressive sensing and direct-binary search algorithm

Abstract

One problem common to imaging techniques based on coherent light is speckle noise. This phenomenon is caused mostly by random interference of light scattered by rough surfaces. Speckle noise can be avoided by using advanced holographic imaging techniques such as optical scanning holography. A more widely known method is to capture several holograms of the same object and to perform an averaging operation so that the signal to noise ratio can be improved. Several digital filters were also proposed to reduce noise in the numerical reconstruction plane of holograms, even though they usually require finding a compromise between noise reduction and edge preservation. In this study, we used a digital filter based on compressive sensing algorithm. This approach enables to obtain results equivalent to the average of multiple holograms, but only a single hologram is needed. Filters for speckle reduction are applied on numerical reconstructions of hologram, and not on the hologram itself. Then, optical reconstruction cannot be performed. We propose a method based on direct-binary search (DBS) algorithm to generate binary holograms that can be reconstructed optically after application of a speckle reduction filter. Since the optimization procedure of the DBS algorithm is performed in the image plane, speckle reduction techniques can be applied on the complex hologram and used as a reference to obtain a binary pattern where the speckle noise generated during the recording of the hologram has been filtered.