Split Bregman Image Synthesis in Radio Interferometry

The Split Bregman (SB) method is a convex optimization algorithm for L1-regularized problems that has been used in denoising, segmentation, reconstruction and other image processing tasks in a compressed sensing setup, but it has received little attention in radio interferometric imaging. It is very efficient and it can be readily implemented in parallel. In this work we apply the SB algorithm and formulate five L1 optimization problems for radio interferometric imaging. We develop an object oriented Python software that facilitates the incorporation a new SB solvers, by inheriting from base classes. We employ two ALMA datasets of protoplanetary disks to test our framework.