Caractérisation de bâtiments par interférométrie SAR polarimétrique multi-images utilisant une méthode d'analyse spectrale adaptative.

This paper proposes a polarimetric weighted subspace fitting spectral analysis method (FP-WSF) to localize and characterize backscattered signals in dense urban environments, by means of a small dimension irregular SAR array in interferometric configuration. Unlike other High Resolution (HR) methods, this technique adapts to the nature of the observed scattering mechanism by considering two signal models, i.e. scattered by coherent and distributed media. An analytical solution to the polarimetric optimization FP-WSF criterion is proposed that permits to maintain a low level of computational complexity. The effectiveness of the proposed method is demonstrated using polarimetric and interferometric multibaseline L-band SAR (MB-POLinSAR) data.