Projection pursuit classification of multiband polarimetric SAR land images

Results are presented for an experiment utilizing a pastoral land scene wit h a variety of eight classes, imaged by the NRL dual band (X and L) polarim etric synthetic aperture radar (NUWSAR) at a spatial resolution of 1.2 in. Projection pursuit statistical analysis tools were applied to a set of simu ltaneous L and X-band fully polarized images (six independent channels) to demonstrate the utility of land classification at high spatial resolution f rom a light aircraft using SAR. The statistical confusion matrix was used a s a quantitative optimization measure of classification. Samples of eight c lasses from a portion of the scene were used to define a training set, then projection pursuit (PP) tools were used for classification. It is clear th at L-band and X-band fully polarized data view the classes in a significant ly different manner, and each brings independent information to the analysi s. These results are not meant to be exhaustive at this time but to demonst rate the utility of applying PP tools to multiband and polarization SAR dat a and to give an indication of the quality of classification one can achiev e with moderately high spatial resolution SAR data using a light plane plat form.