A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Characterization and Correction of Multilook Effects on Eigendecomposition Parameters in PolSAR Images
The eigendecompostion of the coherency matrix and the related parameters, i.e., entropy H, α, and anisotropy A, are effective and popular tools for the analysis and quantitative estimation of the physical parameters of polarimetric synthetic aperture radar images. However, the speckle noise constitutes the main obstacle that hinders these goals and should be filtered. In this paper, based on studies obtained from extensive simulated data sets, we tried to determine how this noise is transmitted to sample eigendecomposition parameters. The dependence between means and variances of sample parameters leads to building their speckle models and to the definition of a bias elimination procedure. We found that sample eigenvalues were affected by multiplicative noise, whereas sample entropy and sample anisotropy were affected by a mixture of multiplicative and additive noise sources. In addition to its versatility, independence to knowledge of the equivalent number of looks and high ability to bias compensation, the proposed bias suppression technique reduced the variance of noise. Simulated and real data as well as the existing theories are used for validation in this paper.
Characterization and Correction of Multilook Effects on Eigendecomposition Parameters in PolSAR Images
The eigendecompostion of the coherency matrix and the related parameters, i.e., entropy H, α, and anisotropy A, are effective and popular tools for the analysis and quantitative estimation of the physical parameters of polarimetric synthetic aperture radar images. However, the speckle noise constitutes the main obstacle that hinders these goals and should be filtered. In this paper, based on studies obtained from extensive simulated data sets, we tried to determine how this noise is transmitted to sample eigendecomposition parameters. The dependence between means and variances of sample parameters leads to building their speckle models and to the definition of a bias elimination procedure. We found that sample eigenvalues were affected by multiplicative noise, whereas sample entropy and sample anisotropy were affected by a mixture of multiplicative and additive noise sources. In addition to its versatility, independence to knowledge of the equivalent number of looks and high ability to bias compensation, the proposed bias suppression technique reduced the variance of noise. Simulated and real data as well as the existing theories are used for validation in this paper.
Characterization and Correction of Multilook Effects on Eigendecomposition Parameters in PolSAR Images
Yahia, Mohamed (author) / Aguili, Taoufik
2015
Article (Journal)
English
Local classification TIB:
770/3710/5670
BKL:
38.03
Methoden und Techniken der Geowissenschaften
/
74.41
Luftaufnahmen, Photogrammetrie
COMMUNICATIONS - On Nonparametric Edge Detection in Multilook SAR Images
| Online Contents | 1998
Data-Level Fusion of Multilook Inverse Synthetic Aperture Radar Images
| Online Contents | 2008
Change detection of multilook polarimetric SAR images using heterogeneous clutter models
| Online Contents | 2014
Probability Density Functions for Multilook Polarimetric Signatures
| Online Contents | 1994