Detail-preserving segmentation of polarimetric SAR imagery

Polarimetric SAR data help researchers understand and quantify the scattering properties of natural scenes, and their variability under different conditions, better than single polarization SAR data. In this paper, we propose two detail-preserving segmentation algorithms for multiband images and evaluate their performance using real multi-frequency multi-polarization SAR complex data. Adaptive neighborhood structures are selected for modeling the polarimetric complex amplitudes and the region labels, and for achieving detail preservation. Experimental results obtained from multi-band and multi-polarization SIR-C data, selected for archaeological applications studies, show that the novel schemes produce significant visual improvements for detail preservation, and exhibit equivalent or higher classification performance with respect to the classical classification schemes.