Graph Based Imaging for Synthetic Aperture Radar

TL;DR

This paper introduces a graph signal processing approach for synthetic aperture radar imaging, enhancing image resolution and denoising while integrating compressed sensing techniques.

Contribution

It presents a novel graph-based imaging method using a modified fused LASSO and ADMM, improving radar image quality over traditional methods.

Findings

Enhanced denoising and resolution in radar images

Effective integration with compressed sensing frameworks

Experimental validation confirms improved image quality

Abstract

In this paper, we propose graph signal processing based imaging for synthetic aperture radar. We present a modified version of fused least absolute shrinkage and selection operator to cater for graph structure of the radar image. We solve the cost function via alternating direction method of multipliers. Our method provides improved denoising and resolution enhancing capabilities. It can also accommodate the compressed sensing framework quite easily. Experimental results corroborate the validity of our proposed methodology.