Phase Unwrapping for InSAR using convex feasible model

In this paper, we proposed a complete scheme for phase retrieval under the framework of convex feasible models. The scheme is composed of two steps: the phase unwrapping and denoising. At first, we proposed a novel phase unwrapping method called Ideal Unwrapper (IU). Afterwards, to solve the rest denoising problem, the theory of convex feasible is introduced. In addition, the convex sets for the Projection Onto Convex Sets (POCs) denoising, which might preserve the given arbitrary feature areas perfectly, are also constructed. Finally, numerical results are also illustrated to show the virtues of the proposed scheme. Especially, the proposed Ideal Unwrapper shows its powerful efficiency in Unwrapping two-dimensional phase images with little computational burden by comparing with the newly proposed Graph Cut method.