<< Click to Display Table of Contents >> Navigation:  Help Content > Gamma and Gaussian Filtering Module > Gamma and Gaussian Filtering Module - Overview

A Note on the Gamma and Gaussian Filtering module

 

This module is aimed at providing a whole family of SAR specific filters, which are particularly efficient to reduce speckle noise, while preserving the radar reflectivity, the textural properties and the spatial resolution, especially in strongly textured SAR images. The algorithms, developed by PRIVATEERS N.V. (Private Experts in Remote Sensing), have been implemented in the most rigorous way, in order to exploit at the best the speckle suppression capabilities.

 

Actually one of the major source of errors in the estimation of soil roughness and soil wetness, which are both crucial parameters for SAR data analysis related to a number of application domains (i.e. agriculture, forestry, snow mapping, etc.), is the presence of speckle. Four Bayesian vector speckle filters have been implemented in order to optimize the speckle removal procedure, depending on the input data type.

 

Note that:

 

–In case of SAR RAW products, the data must be imported and focussed (refer to Focusing module). 

–Default setting for selected parameters can be specified in the Preferences panel.

–The SAR Tutorial, which includes basic knowledge on SAR theory and data processing, complements the online help.

–Data geocoded to GEO-GLOBAL cartographic reference system can be automatically displayed into the Google Earth environment by double clicking on the output .kml file.

–The module has been jointly developed by sarmap s.a. and PRIVATEERS N.V. (Private Experts in Remote Sensing).

 

Technical Note

 

In presence of scene texture, to preserve the useful spatial resolution - for instance to restore the spatial fluctuations of the radar reflectivity (texture) - an A Priori first order statistical model is needed. With regard to SAR clutter, it is well known that the Gamma-distributed scene model is the most appropriate approach in case of data acquired over natural areas (not artificial objects) without too rough morphology. Under such conditions the Gamma-distributed scene model, modulated by either an independent complex-Gaussian speckle model (for SAR Single Look Complex data), or by a Gamma speckle model (for multi-looked SAR Intensity images), gives rise to a K-distributed clutter.

 

Note that the (real) Gaussian distributed scene model remains still popular, mainly for mathematical tractability of the inverse problem in case of multi-channel detected SAR images (multivariate A Priori scene distributions). Nevertheless, in the presence of strong mixed texture or strong relief or artificial object - especially processing Very High Resolution data - it may be hazardous to make an assumption about the statistical distribution of the radar reflectivity under the form of an analytical first order statistical model such as the Gamma distribution; this is why one may wish to make the alternative choice to introduce a Maximum Entropy constraint on texture (“Distribution Entropy”, DE).

 

In this context, the following filter families has been developed:

 

Single Look Complex

-SLC Gaussian DE MAP filter.

-Gaussian Gamma MAP filter.

-Gaussian DE MAP filter.

 

Single Channel Intensity

 

Multi Channel Intensity

 

Polarimetric 

 

References 

 

Nezry E., A. Lopes, and F. Yakam-Simen. "Prior scene knowledge for the Bayesian restoration of mono and multi-channel SAR images". Proceedings of IGARSS’97 Symposium, 1997.

 

E. Nezry and F. Yakam Simen, 1999: "New distribution-entropy Maximum A Posteriori speckle filters for detected, complex, and polarimetric SAR data". Proceedings of IGARSS'99 Symposium, Vol.3, 3 p., Hamburg (Germany), 28 June - 02 July 1999.

 

E. Nezry and F. Yakam Simen, 1999: "A family of distribution-entropy MAP speckle filters for polarimetric SAR data, and for single or multi-channel detected and complex SAR images". Proceedings of the CEOS SAR Workshop, ESA SP-450, pp.219-223, Toulouse (France), 26-29 October 1999.

 

E. Nezry: ''Adaptive Speckle Filtering in Radar Imagery'' from the edited volume Land Applications of Radar Remote Sensin Edited by F. Holecz, P. Pasquali, N. Milisavljevic and D. Closson, 2014.