Abstract
The problem of reconstructing an object from the measurements of the light itscatters is common in numerous imaging applications. While the most popularformulations of the problem are based on linearizing the object-lightrelationship, there is an increased interest in considering nonlinearformulations that can account for multiple light scattering. In this paper, wepropose an image reconstruction method, called CISOR, for nonlinear diffractiveimaging, based on a nonconvex optimization formulation with total variation(TV) regularization. The nonconvex solver used in CISOR is our new variant offast iterative shrinkage/thresholding algorithm (FISTA). We provide fast andmemory-efficient implementation of the new FISTA variant and prove that itreliably converges for our nonconvex optimization problem. In addition, wesystematically compare our method with other state-of-the-art methods onsimulated as well as experimentally measured data in both 2D and 3D settings.