Speckle filtering, segmentation and classification of polarimetric SAR data: a unified approach based on the wavelet transform

A discrete dyadic wavelet transform used as a differential operator was introduced bu S. Mallat et al. (1992) and is a powerful analysis tool to measure the local regularity of a signal. Also an image reconstruction algorithm from the wavelet transform maxima was proven. Based on this theory the present authors propose an approach to map sharp transitions (singularities) and homogeneous areas in SAR imagery. This information in turn lends support in a unified manner to the implementation of speckle filters, segmentation and classification algorithms. Examples of the technique related to polarimetric SAR data processing are reported.