Non-local means SAR despeckling based on scattering

Speckle reduction in Synthetic Aperture Radar (SAR) images is an essential pre-processing step for a correct analysis and interpretation of SAR data. This justifies the huge effort in the image processing community to develop more and more accurate despeckling techniques in order to reduce speckle effects and then improve readability of SAR imagery also for non SAR expert users. Up to now, nonlocal means approaches provide the most promising and effective despeckling performances. In this paper we develop a new non-local means despeckling technique based on electromagnetic scattering mechanisms. The proposed method, based on a physically meaningful similarity criterion for distance evaluation, is theoretically assessed, tested on a simulated SAR image, and compared to the state of the art.

[1]  Luisa Verdoliva,et al.  A Nonlocal SAR Image Denoising Algorithm Based on LLMMSE Wavelet Shrinkage , 2012, IEEE Transactions on Geoscience and Remote Sensing.

[2]  Luisa Verdoliva,et al.  Optical-Driven Nonlocal SAR Despeckling , 2015, IEEE Geoscience and Remote Sensing Letters.

[3]  Florence Tupin,et al.  Iterative Weighted Maximum Likelihood Denoising With Probabilistic Patch-Based Weights , 2009, IEEE Transactions on Image Processing.

[4]  Antonio Iodice,et al.  On shape from shading and SAR images: An overview and a new perspective , 2014, 2014 IEEE Geoscience and Remote Sensing Symposium.

[5]  Luisa Verdoliva,et al.  Benchmarking Framework for SAR Despeckling , 2014, IEEE Transactions on Geoscience and Remote Sensing.

[6]  Giorgio Franceschetti,et al.  SARAS: a synthetic aperture radar (SAR) raw signal simulator , 1992, IEEE Trans. Geosci. Remote. Sens..

[7]  Jean-Michel Morel,et al.  A Review of Image Denoising Algorithms, with a New One , 2005, Multiscale Model. Simul..