Recovery of Partially Corrupted SAR Images by Super-Resolution Based on Spectrum Extrapolation

Abstract

The problem of chirped synthetic aperture radar (SAR) systems is the high vulnerability of the received information to electromagnetic (EM) attacks. This letter proposes a valid recovery solution for SAR single-look complex images that are corrupted by noncoherent EM noise covering only the higher frequency spectrum. The solution consists of, first, exporting the spectrum damages that occur in the native data and, second, focusing only the survived spectrum information at lower resolution. The recovery of the original image is done by super-resolution signal processing based on spectrum extrapolation and implemented by convex programming.