Analysis of SAR Response Anisotropic Behavior Using Sub-Aperture Polarimetric Data
University of Rennes 1,
Campus de Beaulieu - Bat 11.D,
35042 Rennes cedex,
(2) University of Illinois at Chicago, 900 W. Taylor St., SEL (607)W-4210,, CHICAGO, IL, 60607-7018, United States
(3) Technical University of Berlin, Straße des 17. Juni 135, EB9, D-10623 Berlin, Germany
In synthetic aperture radar (SAR) polarimetry, the measured polarimetric signatures are used to analyse physical scattering properties of the imaged objects. It is thereby gernerally assumed, that the sensor has a fixed orientation with respect to the objects. But SAR sensors operating at lower frequencies, like L- and P-band, have a wide antenna characteristic in azimuth; i.e. during the formation of the synthetic aperture, multiple squint angles are integrated to form the full-resolution SAR image. Possible changes in the polarimetric backscattering characteristics during this change of the look-angle remain unconsidered. In this paper a fully polarimetric sub-aperture analysis method is introduced. Using deconvolution, synthesized SAR images are decomposed into sub-aperture data sets, which correspond to the scene responses under different azimuthal look angles. A statistical analysis of the polarimetric data permits to clearly discriminate the media showing a varying behavior during the SAR integration. This analysis is based on test staristics applied to R sub-aperture images and uses the Wishart pdf of a coherency matrix. This test permits to disrciminate, for each pixel, the range of azimuth observation angles corresponding to an anisotropic behavior. Some plowed surfaces show, for some particular observatoon angle, a resonant behavior, which has a significant influence on pertinent polarimetric indicators. Certain surface parameters of such areas may be estimated, using the subaperture approach, which are not observable in the full-resolution data. Finally, a method is proposed in this paper, which eliminates the influence of azimutal backscattering variations in conventional polarimetric SAR data analysis. The effectivness of the new methods is demonstrated on fully polarimetric SAR data, acquired by the DLR s airborne experimental SAR sensor (E-SAR) in L-band.