Ship Detection Using Compact Polarimetric SAR Based on the Notch Filter

Abstract

Compact polarimetric data exploitation, especially in hybrid-polarimetric (HP) mode, is currently attracting increasing interest due to the new generation of synthetic aperture radar (SAR) systems. Recently, it has been demonstrated that the notch filter is useful for ship detection in either full- or dual-polarization (DP)-mode SAR images. In this paper, the notch filter investigation is further extended to HP SAR architecture for ship detection on the ocean surface. First, a version of the notch filter that is suitable for HP SAR is proposed based on the definition of the corresponding feature partial scattering vector from the covariance matrix of the HP SAR. Subsequently, a novel model characterizing the statistics of the notch distance of sea clutter in the HP mode is developed. Based on the statistical model, the threshold of constant false-alarm rate (CFAR) ship detection is theoretically and analytically derived, which allows the automatic and adaptive implementation for ship detection in varying sea backgrounds in practical applications. Experiments on the HP SAR data emulated from full-polarization L-band Aerospace Exploration Agency Advanced Land Observation Satellite Phased-Array type L-band SAR and C-band RADARSAT-2 SAR measurements validate not only the soundness of the proposed CFAR detection but also the high accuracy of the presented model in fitting HP SAR data. Furthermore, the notch filter and its CFAR realization provide the same benchmark for the comparison of the detectability of HP and conventional linear DP SAR data. Preliminary findings suggest that the detection performance of HP SAR is superior to that of DP SAR in ship observation. Therefore, the proposed CFAR method based on the notch filter provides a promising technique for the detection of ships using HP SAR data.