Wilcoxon Nonparametric CFAR Scheme for Ship Detection in SAR Image

Abstract

The parametric constant false alarm rate (CFAR) detection algorithms, which are based on various statistical distributions, such as Gaussian, Gamma, Weibull, log-normal, G0, and alpha-stable distribution, are most widely used to detect the ship targets in SAR images at present. However, the clutter background in SAR images is complicated and variable. When the actual clutter background deviates from the assumed statistical distribution, the performance of the parametric CFAR detector deteriorates, whereas the advantage of the nonparametric CFAR detector that its false alarm rate is independent of the background distribution is exhibited. In this work, the Wilcoxon nonparametric CFAR scheme for the ship detection in SAR images is proposed and analyzed, and a closed form of the false alarm rate for the Wilcoxon nonparametric CFAR detector to determine the decision threshold is presented. By comparison with several typical parametric CFAR schemes on Sentinel-1A, ICEYE-X6, and Gaofen-3 SAR images, the robustness of the ability of the Wilcoxon nonparametric CFAR detector to control the actual false alarm rate at a suitably low level in different detection backgrounds is revealed, and its detection performance for the weak ships in the rough sea backgrounds is evidently improved. Moreover, the detection speed of the Wilcoxon nonparametric CFAR detector is fast, and it has a simple hardware implementation.