Enhanced target extraction in strong clutter scene

Abstract

Clutter suppression is always problematic in synthetic aperture radar (SAR) imaging. Especially when clutter is strong, inappropriate threshold on image amplitude may lead to failure to completely separate the background and the region of interest (ROI). Therefore, this paper firstly presents an adaptive threshold estimation method from the view of coherence. To protect weak scatterers, neighbored points around the extracted centers are also preserved. With the strong sparsity of the ROI in the image, the resolution of the extracted target region is enhanced by the sparse signal processing technique. Experimental results on simulated and real data both confirm its practicability.