SAR Image Scattering Center Estimation via Atomic Norm Minimization

Scatter center matching is one of the choke points of radar automatic target recognition (RATR) based on Synthetic Aperture Radar (SAR). Accuracy of scattering center extraction is an important factor affecting target recognition performance. This paper proposes a SAR images scattering center estimation method based on atomic norm minimization (ANM) model. This method models the scattering center estimation problem as a two-dimensional line spectrum estimation problem, then the estimation is performed in a continuous domain via ANM. The experimental results indicate the superiorities of the proposed method in terms of accuracy and robustness.

I. Introduction

SAR image plays a significant role in radar automatic target recognition (RATR) [1]. Scatter center matching is one of the choke points of RATR based on Synthetic Aperture Radar (SAR) [2]. In the optical region, the electromagnetic scattering of a target can be approximately described by the scattering center model. Furthermore, scattering center model is a key component of RATR based on SAR image [3]. Therefore, accurately estimating the location and the amplitude of these scattering centers is a crucial step in SAR RATR, which is mainly discussed in this article [4] [5].