Iterative focus detection in hologram tomography.

Hologram tomography is a two-step method for three-dimensional topometry of extended objects. The first step consists of the hologram recording with a single laser pulse of 35 ns duration and storage in a photosensitive material. In the second step the hologram is optically reconstructed and digitized, which leads to a set of two-dimensional projections at different axial positions. A maximization of a focus measure has to be performed to extract the surface position out of the projections. Unlike with well-established methods, where the comparison of sharpness values is done parallel to the optical axis, we propose an iterative solution to perform the maximization along the direction of image formation, which is evaluated for each surface point individually. This leads to a better reproducibility of the surface in the off-axis regions.