Guillaso, Stéphane; Jäger, Marc; Reigber, Andreas (Berlin University of Technology, Berlin, Germany)
This paper describes the extraction of information from polarimetric tomographic SAR data. This analysis is performed in three steps. Polarimetric tomographic data are first processed using a Capon-based vectorial spectral analysis technique, in order to extract physical and geometrical information over vertical structures. Then, a feature detection procedure with automatic scale selection is applied to the obtained reflectivity tomograms, making possible to localise regions of interest. Finally, a coherent polarimetric decomposition, the Cameron-based decompostion, is used to physically characterise the detected scatterer.