PolSAR – New Methodologies for Polarimetric SAR Image Analysis
Synthetic Aperture Radar is a common tool in today’s remote sensing. Many modern SAR systems measure also the target’s backscattering dependence on SAR signal polarization. This allows distinguishing between different reflection types as SAR signal polarization usually changes during the reflection. SAR imaging and interpretation techniques which utilize polarization changes and the related scientific discipline are called polarimetry. Polarimetry is used mainly to improve image classification and target detection.
Aalto University has been working with polarimetric techniques and image analysis since 1996, when first fully polarimetric images in the ESA EMAC-95 campaign were acquired. Main focus areas have been polarimetric decompositions, image interpretation and optimization of image parameterization.
Figure 1. A polarimetric AIRSAR image of San Francisco, presented by using multiresolution synoptic visualization scheme and polarimetric parameters developed in the project.
Cooperation partners: ONERA, DLR
Recent publications:
- J. Praks, E. Colin Koeniguer, M. Hallikainen, ”Alternatives to target entropy and alpha angle in SAR Polarimetry,” IEEE Transactions on Geoscience and Remote Sensing, vol. 47, no. 7, pp. 2262–2274, July 2009.
- J. Praks, M. Hallikainen, and E.C. Koeniguer, ”Polarimetric SAR image visualization and interpretation with covariance matrix invariants,” Proc. IEEE Geoscience and Remote Sensing Symposium (IGARSS’10), pp. 2035–2038, Honolulu, USA, 25-30 July 2010.
- E. Colin-Koeniguer, N. N. Trouvé, and J. Praks, “A review about alternatives to classical polarimetric SAR parameters,” Proc. EUSAR, 8th European Conference on Synthetic Aperture Radar, Aachen, Germany, 7-10 June 2010.