Analysis of Adaptive and Advanced Speckle Filters on SAR Data
Arundhati Ray Misra, Dhwani Ajmera · IOSR Journal of Computer Engineering · 2017
Synthetic Aperture RADAR(SAR) images get inherently affected by speckle noise which is multiplicative in nature.This noise affects the image spatial statistics and properties.Over the past several years, many SAR denoising algorithms have been developed to reduce speckle noise.Some of the standard speckle filters are Gamma MAP, Lee, Frost and Kuan filters.Further, these have also been modified to obtain better results after filtering, than their original counterparts.Apart from the standard speckle filters, advanced SAR filters like Block Matching 3 Dimensional (BM3D) are also present.In this paper several standard as well as advanced speckle filters have been analyzed and compared.For comparison, Quality Assessment has been performed where the filtered images are compared to each other using parameters like Radiometric Resolution and others.These parameters help to distinguish the performance of the filters on basis of signal strength, speckle reduction, mean preservation and edge and feature preservation.In the paper, radiometric resolution, speckle index and mean preservation index will be used to analyze among the performance of the filters.