SAR image denoising: A multiscale robust statistical approach

A. Achim, A. Bezerianos, P. Tsakalides

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Scopus citations


Synthetic aperture radar (SAR) images are inherently affected by multiplicative speckle noise, which is due to the coherent nature of the scattering phenomenon. It appears sensible to reduce speckle in SAR images, provided that the structural features and textural information are not lost. We present a novel speckle removal algorithm within the framework of wavelet analysis. First, we show that the subband decompositions of logarithmically transformed SAR images are best described by alpha-stable distributions, a family of heavy-tailed densities. Consequently, we design a maximum a posteriori (MAP) estimator that exploits this a priori information. We use the alpha-stable model to develop a blind speckle-suppression processor that performs a non-linear operation on the data, and we relate this non-linearity to the degree of non-Gaussianity of the data. Finally, we compare our proposed method to a current state-of-the-art soft thresholding technique applied on an aerial image and we quantify the achieved performance improvement.

Original languageEnglish (US)
Title of host publication2002 14th International Conference on Digital Signal Processing Proceedings, DSP 2002
EditorsA.N. Skodras, A.G. Constantinides
PublisherInstitute of Electrical and Electronics Engineers Inc.
Number of pages4
ISBN (Electronic)0780375033
StatePublished - 2002
Event14th International Conference on Digital Signal Processing, DSP 2002 - Santorini, Hellas, Greece
Duration: Jul 1 2002Jul 3 2002

Publication series

NameInternational Conference on Digital Signal Processing, DSP


Other14th International Conference on Digital Signal Processing, DSP 2002
CitySantorini, Hellas

ASJC Scopus subject areas

  • Signal Processing


Dive into the research topics of 'SAR image denoising: A multiscale robust statistical approach'. Together they form a unique fingerprint.

Cite this