TomoSAR Imaging Using Statistical Regularization on Polarimetric SAR Observations

Abstract

Aimed at estimating the position of the vertical structures that scatter the field back towards the active sensor, synthetic aperture radar (SAR) tomography (TomoSAR) focusing techniques (e.g., beamforming, Capon and multiple signal classification) determine how energy distributes over space. Furthermore, their polarimetric configurations pursue finding optimal polarization combinations in order to recover the polarimetric pseudo-power and extract the associated scattering mechanisms and height of reflectors, identified through the highest local maxima. Seeking to attain finer resolution, this article analyses the usage of statistical regularization on polarimetric SAR observations for TomoSAR imaging. The retrievals of conventional polarimetric TomoSAR focusing methods are to be refined in a post-processing step, making use of weighted covariance fitting. The main goal of this article is introducing the corresponding algorithm and pointing out, via simulations, the main advantages and drawbacks of the proposed novel strategy.