Enhancements of an Adaptive Neighborhood Speckle Filtering Algorithm to Improve Analysis of Polarimetric SAR Imagery

Image noise often hinders the analysis and understanding of most remote sensing imagery. Synthetic aperture radar (SAR) images, in particular, suffer from a phenomenon known as speckle. Speckle is the result of the coherent sum of scattering mechanisms within a resolution cell and is typically modeled as multiplicative noise. Numerous approaches have been taken to reduce speckle, the most promising of which are adaptive and statistically based. These algorithms typically take advantage of the statistics of speckle or the multi-scale image structure to detect regions of similarity. Several of these algorithms also take advantage of the multi-channel nature of polarimetric SAR to improve the region detection. This work presents two enhancements to such an algorithm with an application toward polarimetric parameter estimation and change detection and presents results using the Japanese satellite-based Phased Array L-band Synthetic Aperture Radar (PALSAR) system.