Polarimetric SAR imaging for extended targets in clutter

This paper presents an analytic inversion scheme for polarimetric synthetic-aperture radar in the case of an extended target embedded in clutter and thermal noise. We developed a filtered-backprojection-type reconstruction method. We model targets and clutter as collections of dipoles, so that in each pixel we reconstruct a scattering matrix. In addition, we implement directional scattering assumptions for the curve-like extended target. For the inversion we choose to find the filter which minimizes the mean-square error between the reconstructed image and the actual target scattering matrix. Our work differs from standard polarimetric SAR imaging in that we do not perform channel-by-channel image reconstruction. We find that it is preferable to use a coupled reconstruction scheme in which we use all sets of collected data to form each element of the scattering matrix. We show in our numerical experiments that the coupled technique not only minimizes the mean-square error, but also increases the image signal-to-clutter ratio.