A segmentation-based CFAR algorithm for subsurface targets detection in FLGPSAR

Forward-Looking Ground Penetrating Synthetic Aperture Radar (FLGPSAR) has the capability of forming two-dimensional high-resolution images of subsurface objects from a standoff distance. This paper addresses the detection of subsurface targets, i.e. landmines, in FLGPSAR images. The conventional Constant False-Alarm Rate (CFAR) algorithm has been widely used in SAR image target detection, but its performance will degrade in subsurface targets detection because of the presence of interfering targets and clutter power transition. In this paper, a segmentation-based CFAR (S-CFAR) algorithm is proposed. The S-CFAR algorithm can remove the interfering targets or clutter power transition before estimating the parameters of the clutter background to achieve a better performance than the conventional CFAR algorithm. The real data processing results are given to validate the efficiency of the proposed method.