A comparative sensitivity analysis of scattering-based despeckling algorithms

Synthetic aperture radar (SAR) image interpretation and analysis are heavily affected by the speckle noise typical of coherent acquisition systems. Very recently, a novel despeckling approach based on the exploitation of a priori information about the scattering behavior of the illuminated surface has been presented. The new scattering-based algorithms, named SB-PPB and SB-SARBM3D, represent a scattering-oriented version of the preexisting PPB and SARBM3D algorithms. In this paper, a comparative sensitivity analysis of the abovementioned despeckling algorithms is performed by means of SAR image simulation tools. First, the sensitivity of the filters against the use of different electromagnetic scattering models for the illuminated surface is analyzed. Then, the roles of the digital elevation model resolution and of the coregistration step are also investigated.

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

[2]  Luisa Verdoliva,et al.  Scattering-Based SARBM3D , 2016, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[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.  Non-local means SAR despeckling based on scattering , 2015, 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).

[5]  Antonio Iodice,et al.  Scattering-Based Nonlocal Means SAR Despeckling , 2016, IEEE Transactions on Geoscience and Remote Sensing.

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