Phaseless Passive Synthetic Aperture Imaging with Regularized Wirtinger Flow

In this paper, we present a training-less methodology for Phaseless Passive Synthetic Aperture Radar imaging. The existing approach based on Wirtinger Flow (WF) requires large number of phaseless measurements for satisfactory reconstruction. To address this issue, we propose a regularized Wirtinger Flow based approach that helps with efficient image reconstruction. We employ Total Variation, BM3D and Deep Image Prior based regularizers/denoisers in an ADMM framework for the proposed solution. The results indicate that compared to the state-of-the-art, the proposed approach not only facilitates better reconstruction with lesser measurements but also shows robustness against SAR trajectory errors.