Ship Detection and Characterization Using Polarimetric SAR Data

Abstract

Synthetic Aperture Radar (SAR) sensors offer the coverage rate and all-weather capability necessary for wide-area surveillance and can provide very useful information for the detection and characterization of surface targets. Conventional SARs operate with a single polarization channel, while recent and future spaceborne SARs (Envisat ASAR, Radarsar-2) will offer the possibility to use multiple polarization channels, which will enable better detection and characterization of man-made objects. Standard target detection approaches on SAR images consist in the application of a Cell Averaging or 2-parameters CFAR detector and usually produce a large number of false alarms. This large number of false alarms prohibits their manual rejection. However, over the past ten years a number of algorithms have been proposed to extract information from a polarimetric SAR scattering matrix in order to enhance and/or characterize man-made objects. Target decomposition algorithms, such as Cameron’s coherent target decomposition (CTD) and the Odd-Even basis decomposition, provide information for detecting ships and reducing false alarms. On the other hand, information derived from spectral analysis, such as the subaperture coherence, enhance the ship’s signature and allow a better discrimination of the desired target. The evidential fusion of such information can lead to the automatic rejection of the false alarms generated by the CFAR detector. In addition, the aforementioned information can lead to a better characterization of detected targets. This paper presents an approach, as well as results, for automatic target detection using the evidential fusion of polarimetric features and the spectral analysis technique. Preliminary results for ship characterization using polarimetric information are also presented. This research is motivated by airborne/spaceborne surveillance applications such as land and coastal surveillance missions using SAR/PolSAR imagery.