Time-frequency analysis of polarimetric SAR images

This article demonstrates the usefulness of time-frequency tools in SAR imagery through two applications. A new method is proposed to analyse moving targets. It uses a two dimensional matching-pursuit algorithm, over a time-frequency dictionary. Point targets speed and scattering characteristics (frequency comportment, angular and polarimetric behaviour) can be retrieved, thus permitting a better knowledge of the target. Urban areas analysis is carried on in a time-frequency framework. The behaviour of pixels in the frequency domain is studied and employed as an input for two new indicators, allowing a classification of urban areas pixels