One-dimensional surface profile retrieval from grazing incidence images under coherent X-ray illumination

Three different approaches to the problem of mirror surface retrieval by means of coherent X-ray imaging are analyzed. First, an analytical approximation to the problem is presented. Then, an iterative algorithm is applied to the linearized problem. Finally an algorithm based on artificial neural networks is advanced. To evaluate the performance of the algorithms described, noise-free simulations with synthetically generated data were performed. It is demonstrated that the examined algorithms can be combined with each other to improve the overall performance, and can be used for data analysis in X-ray-based mirror metrology.