Classification of SAR images based on estimates of the parameters of the gA0 distribution

There are many statistical models for Synthetic Aperture Radar (SAR) images. Among them, the multiplicative model is based on the assumption that the observed random field Z is the result of the product of two independent and unobserved random fields: X and Y. The random field X models the backscatt...

Descripción completa

Guardado en:
Detalles Bibliográficos
Publicado: 1998
Materias:
Asymptotic stability
Backscattering
Convergence of numerical methods
Mathematical models
Maximum likelihood estimation
Parameter estimation
Probability density function
Probability distributions
Radar imaging
Speckle
Statistical methods
Synthetic aperture radar
Image classification
Multiplicative model
Statistical models
Supervised classification
Unsupervised classification
Image analysis
Acceso en línea:https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_0277786X_v3457_n_p202_Chomczimsky
http://hdl.handle.net/20.500.12110/paper_0277786X_v3457_n_p202_Chomczimsky
Aporte de:
Biblioteca Digital - Facultad de Ciencias Exactas y Naturales (UBA) de Universidad de Buenos Aires
id paper:paper_0277786X_v3457_n_p202_Chomczimsky
record_format dspace
spelling paper:paper_0277786X_v3457_n_p202_Chomczimsky2023-06-08T15:26:10Z Classification of SAR images based on estimates of the parameters of the gA0 distribution Asymptotic stability Backscattering Convergence of numerical methods Mathematical models Maximum likelihood estimation Parameter estimation Probability density function Probability distributions Radar imaging Speckle Statistical methods Synthetic aperture radar Image classification Multiplicative model Statistical models Supervised classification Unsupervised classification Image analysis There are many statistical models for Synthetic Aperture Radar (SAR) images. Among them, the multiplicative model is based on the assumption that the observed random field Z is the result of the product of two independent and unobserved random fields: X and Y. The random field X models the backscatter, and thus depends only on the type of area each pixel belongs to. On the other hand, the random field Y takes into account that SAR images are the result of a coherent imaging system that produces the well known phenomenon called speckle, and that they are generated by performing an average of n statistically independent images -looks- in order to reduce the speckle effect. There are various ways of modelling the random fields X and Y. Recently Frery et.al, proposed the distributions Γ-1/2 (α, γ) and Γ1/2 (n, n) for of X and Y respectively. This resulted in a new distribution for Z: the gA0 (α, γ, n) distribution. Here, the parameters α and γ depend on the ground truth of each pixel and the parameter n is the number of looks used to generate the image. The advantage of this distribution over the ones used in the past is that it models very well extremely heterogeneous areas like cities, as well as moderately heterogeneous areas like forests, and homogeneous areas like pastures. As the ground truth can be characterized by the parameters α and γ, their estimation for each pixel generates parameter maps that can be used as the input for classical classification methods. In this work, different parameter estimation procedures are used and compared on synthetic and real SAR images, and then, supervised and unsupervised classifications are performed and evaluated. 1998 https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_0277786X_v3457_n_p202_Chomczimsky http://hdl.handle.net/20.500.12110/paper_0277786X_v3457_n_p202_Chomczimsky
institution Universidad de Buenos Aires
institution_str I-28
repository_str R-134
collection Biblioteca Digital - Facultad de Ciencias Exactas y Naturales (UBA)
topic Asymptotic stability
Backscattering
Convergence of numerical methods
Mathematical models
Maximum likelihood estimation
Parameter estimation
Probability density function
Probability distributions
Radar imaging
Speckle
Statistical methods
Synthetic aperture radar
Image classification
Multiplicative model
Statistical models
Supervised classification
Unsupervised classification
Image analysis
spellingShingle Asymptotic stability
Backscattering
Convergence of numerical methods
Mathematical models
Maximum likelihood estimation
Parameter estimation
Probability density function
Probability distributions
Radar imaging
Speckle
Statistical methods
Synthetic aperture radar
Image classification
Multiplicative model
Statistical models
Supervised classification
Unsupervised classification
Image analysis
Classification of SAR images based on estimates of the parameters of the gA0 distribution
topic_facet Asymptotic stability
Backscattering
Convergence of numerical methods
Mathematical models
Maximum likelihood estimation
Parameter estimation
Probability density function
Probability distributions
Radar imaging
Speckle
Statistical methods
Synthetic aperture radar
Image classification
Multiplicative model
Statistical models
Supervised classification
Unsupervised classification
Image analysis
description There are many statistical models for Synthetic Aperture Radar (SAR) images. Among them, the multiplicative model is based on the assumption that the observed random field Z is the result of the product of two independent and unobserved random fields: X and Y. The random field X models the backscatter, and thus depends only on the type of area each pixel belongs to. On the other hand, the random field Y takes into account that SAR images are the result of a coherent imaging system that produces the well known phenomenon called speckle, and that they are generated by performing an average of n statistically independent images -looks- in order to reduce the speckle effect. There are various ways of modelling the random fields X and Y. Recently Frery et.al, proposed the distributions Γ-1/2 (α, γ) and Γ1/2 (n, n) for of X and Y respectively. This resulted in a new distribution for Z: the gA0 (α, γ, n) distribution. Here, the parameters α and γ depend on the ground truth of each pixel and the parameter n is the number of looks used to generate the image. The advantage of this distribution over the ones used in the past is that it models very well extremely heterogeneous areas like cities, as well as moderately heterogeneous areas like forests, and homogeneous areas like pastures. As the ground truth can be characterized by the parameters α and γ, their estimation for each pixel generates parameter maps that can be used as the input for classical classification methods. In this work, different parameter estimation procedures are used and compared on synthetic and real SAR images, and then, supervised and unsupervised classifications are performed and evaluated.
title Classification of SAR images based on estimates of the parameters of the gA0 distribution
title_short Classification of SAR images based on estimates of the parameters of the gA0 distribution
title_full Classification of SAR images based on estimates of the parameters of the gA0 distribution
title_fullStr Classification of SAR images based on estimates of the parameters of the gA0 distribution
title_full_unstemmed Classification of SAR images based on estimates of the parameters of the gA0 distribution
title_sort classification of sar images based on estimates of the parameters of the ga0 distribution
publishDate 1998
url https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_0277786X_v3457_n_p202_Chomczimsky
http://hdl.handle.net/20.500.12110/paper_0277786X_v3457_n_p202_Chomczimsky
_version_ 1768545229928923136

Ejemplares similares