Remote sensing of urban areas from polarimetric SAR data using Time-Frequency and spectral analysis methods

Abstract

This article presents two complementary approaches for the study of urban areas using polarimetric and interferometric SAR (POL-inSAR) data. A multidimensional linear Time-Frequency (TF) decomposition is used to analyze the intrinsic polarimetric behavior of different components of an urban area. A TF signal model, adapted to the case of urban areas, is proposed and studied using relevant statistical descriptors. A TF classification procedure is introduced to retrieve building location and characterize their polarimetric response, and applied to fully polarimetric SAR data acquired by the E-SAR sensor at L-band. Multiple POL-inSAR signals, acquired from different positions, are used to estimate the height of buildings using the interferometry principle. High-performance array signal processing techniques are adapted to the case multi-baseline POL-InSAR (MBPI) observations, in order to enhance the height estimation of scatterers by calculating optimal polarization combinations and determining their scattering characteristics. These multidimensional spectral estimation methods are shown to resolve the building layover problem by extracting and analyzing two components within one azimuth-range resolution cell.