High-fidelity in-line digital holography by hybrid constraints phase retrieval algorithm

Abstract

In-line digital holography (DH) achieves indirect detection of optical complex-amplitudes. Single-shot in-line DH via phase retrieval can obtain reconstructed complex field without twin image by exploring signal priors. However, these iterative projection methods based on single prior suffer from poor imaging fidelity. In this paper, we propose a novel hybrid constraints phase retrieval (HC-PR) algorithm to reconstruct the in-line digital hologram. In particular, the combined methods of the positive absorption, the total variation (TV) minimization, and the denoising regularization act as the prior constraints of the optimization framework to ensure the truth and robustness of the reconstructed complex-amplitude field. Moreover, distinguishing the signal region with the no-signal region enables high fidelity imaging of fractional hologram. The feasibility of our method is verified by the numerical simulation and the optical experiments. The HC-PR exhibits better reconstructed results compared with other popular methods, especially facing the fractional holograms. This new in-line digital holographic imaging approach provides a simple and efficient solution for the design of lensless imaging.