Iterative algorithms for twin-image elimination in in-line holography using finite-support constraints

The quality of reconstructed images from in-line holograms can be seriously degraded by the linear superposition of twin images having the same information but different foci. Starting from the reconstructed field at the real image plane, we make use of the uncontaminated information contained in the out-of-focus wave (virtual image) outside the in-focus wave (real image) support, together with a finite-support constraint, to form an iterative procedure for twin-image elimination. This algorithm can reconstruct complex objects, provided that they are not recorded in very near-field conditions. For real objects additional constraints can be imposed, extending the algorithm application to very near-field conditions. The algorithm’s convergence properties are studied in both cases, and some examples are shown.