Unsupervised classification algorithm based on EM method for polarimetric SAR images

Autores
Fernández Michelli, Juan Ignacio; Hurtado, Martin; Areta, Javier Alberto; Muravchik, Carlos Horacio
Año de publicación
2016
Idioma
inglés
Tipo de recurso
artículo
Estado
versión publicada
Descripción
In this work we develop an iterative classification algorithm using complex Gaussian mixture models for the polarimetric complex SAR data. It is a non supervised algorithm which does not require training data or an initial set of classes. Additionally, it determines the model order from data, which allows representing data structure with minimum complexity. The algorithm consists of four steps: initialization, model selection, refinement and smoothing. After a simple initialization stage, the EM algorithm is iteratively applied in the model selection step to compute the model order and an initial classification for the refinement step. The refinement step uses Classification EM (CEM) to reach the final classification and the smoothing stage improves the results by means of non-linear filtering. The algorithm is applied to both simulated and real Single Look Complex data of the EMISAR mission and compared with the Wishart classification method. We use confusion matrix and kappa statistic to make the comparison for simulated data whose ground-truth is known. We apply Davies-Bouldin index to compare both classifications for real data. The results obtained for both types of data validate our algorithm and show that its performance is comparable to Wishart's in terms of classification quality.
Fil: Fernández Michelli, Juan Ignacio. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales. Universidad Nacional de La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales; Argentina
Fil: Hurtado, Martin. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales. Universidad Nacional de La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales; Argentina
Fil: Areta, Javier Alberto. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional de Río Negro; Argentina
Fil: Muravchik, Carlos Horacio. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales. Universidad Nacional de La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales; Argentina. Universidad Nacional de la Plata. Facultad de Ingeniería. Departamento de Electrotecnia. Laboratorio de Electrónica Industrial, Control e Instrumentación; Argentina. Provincia de Buenos Aires. Gobernación. Comisión de Investigaciones Científicas; Argentina
Materia
Bic
Classification
Expectation Maximization
Gaussian Mixture
Mixture Reduction
Sar Images
Nivel de accesibilidad
acceso abierto
Condiciones de uso
https://creativecommons.org/licenses/by-nc-nd/2.5/ar/
Repositorio
CONICET Digital (CONICET)
Institución
Consejo Nacional de Investigaciones Científicas y Técnicas
OAI Identificador
oai:ri.conicet.gov.ar:11336/24748
Acceder
id CONICETDig_1057fa0b3b5f42f489700a8988192735
oai_identifier_str oai:ri.conicet.gov.ar:11336/24748
network_acronym_str CONICETDig
repository_id_str 3498
network_name_str CONICET Digital (CONICET)
spelling Unsupervised classification algorithm based on EM method for polarimetric SAR imagesFernández Michelli, Juan IgnacioHurtado, MartinAreta, Javier AlbertoMuravchik, Carlos HoracioBicClassificationExpectation MaximizationGaussian MixtureMixture ReductionSar Imageshttps://purl.org/becyt/ford/2.2https://purl.org/becyt/ford/2In this work we develop an iterative classification algorithm using complex Gaussian mixture models for the polarimetric complex SAR data. It is a non supervised algorithm which does not require training data or an initial set of classes. Additionally, it determines the model order from data, which allows representing data structure with minimum complexity. The algorithm consists of four steps: initialization, model selection, refinement and smoothing. After a simple initialization stage, the EM algorithm is iteratively applied in the model selection step to compute the model order and an initial classification for the refinement step. The refinement step uses Classification EM (CEM) to reach the final classification and the smoothing stage improves the results by means of non-linear filtering. The algorithm is applied to both simulated and real Single Look Complex data of the EMISAR mission and compared with the Wishart classification method. We use confusion matrix and kappa statistic to make the comparison for simulated data whose ground-truth is known. We apply Davies-Bouldin index to compare both classifications for real data. The results obtained for both types of data validate our algorithm and show that its performance is comparable to Wishart's in terms of classification quality.Fil: Fernández Michelli, Juan Ignacio. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales. Universidad Nacional de La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales; ArgentinaFil: Hurtado, Martin. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales. Universidad Nacional de La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales; ArgentinaFil: Areta, Javier Alberto. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional de Río Negro; ArgentinaFil: Muravchik, Carlos Horacio. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales. Universidad Nacional de La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales; Argentina. Universidad Nacional de la Plata. Facultad de Ingeniería. Departamento de Electrotecnia. Laboratorio de Electrónica Industrial, Control e Instrumentación; Argentina. Provincia de Buenos Aires. Gobernación. Comisión de Investigaciones Científicas; ArgentinaElsevier Science2016-07info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionhttp://purl.org/coar/resource_type/c_6501info:ar-repo/semantics/articuloapplication/pdfapplication/pdfapplication/pdfapplication/pdfapplication/pdfhttp://hdl.handle.net/11336/24748Fernández Michelli, Juan Ignacio; Hurtado, Martin; Areta, Javier Alberto; Muravchik, Carlos Horacio; Unsupervised classification algorithm based on EM method for polarimetric SAR images; Elsevier Science; Isprs Journal Of Photogrammetry And Remote Sensing; 117; 7-2016; 56-650924-2716CONICET DigitalCONICETenginfo:eu-repo/semantics/altIdentifier/url/http://www.sciencedirect.com/science/article/pii/S0924271616000587info:eu-repo/semantics/altIdentifier/doi/10.1016/j.isprsjprs.2016.03.001info:eu-repo/semantics/openAccesshttps://creativecommons.org/licenses/by-nc-nd/2.5/ar/reponame:CONICET Digital (CONICET)instname:Consejo Nacional de Investigaciones Científicas y Técnicas2025-09-03T09:45:01Zoai:ri.conicet.gov.ar:11336/24748instacron:CONICETInstitucionalhttp://ri.conicet.gov.ar/Organismo científico-tecnológicoNo correspondehttp://ri.conicet.gov.ar/oai/requestdasensio@conicet.gov.ar; lcarlino@conicet.gov.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:34982025-09-03 09:45:01.856CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicasfalse
dc.title.none.fl_str_mv Unsupervised classification algorithm based on EM method for polarimetric SAR images
title Unsupervised classification algorithm based on EM method for polarimetric SAR images
spellingShingle Unsupervised classification algorithm based on EM method for polarimetric SAR images
Fernández Michelli, Juan Ignacio
Bic
Classification
Expectation Maximization
Gaussian Mixture
Mixture Reduction
Sar Images
title_short Unsupervised classification algorithm based on EM method for polarimetric SAR images
title_full Unsupervised classification algorithm based on EM method for polarimetric SAR images
title_fullStr Unsupervised classification algorithm based on EM method for polarimetric SAR images
title_full_unstemmed Unsupervised classification algorithm based on EM method for polarimetric SAR images
title_sort Unsupervised classification algorithm based on EM method for polarimetric SAR images
dc.creator.none.fl_str_mv Fernández Michelli, Juan Ignacio
Hurtado, Martin
Areta, Javier Alberto
Muravchik, Carlos Horacio
author Fernández Michelli, Juan Ignacio
author_facet Fernández Michelli, Juan Ignacio
Hurtado, Martin
Areta, Javier Alberto
Muravchik, Carlos Horacio
author_role author
author2 Hurtado, Martin
Areta, Javier Alberto
Muravchik, Carlos Horacio
author2_role author
author
author
dc.subject.none.fl_str_mv Bic
Classification
Expectation Maximization
Gaussian Mixture
Mixture Reduction
Sar Images
topic Bic
Classification
Expectation Maximization
Gaussian Mixture
Mixture Reduction
Sar Images
purl_subject.fl_str_mv https://purl.org/becyt/ford/2.2
https://purl.org/becyt/ford/2
dc.description.none.fl_txt_mv In this work we develop an iterative classification algorithm using complex Gaussian mixture models for the polarimetric complex SAR data. It is a non supervised algorithm which does not require training data or an initial set of classes. Additionally, it determines the model order from data, which allows representing data structure with minimum complexity. The algorithm consists of four steps: initialization, model selection, refinement and smoothing. After a simple initialization stage, the EM algorithm is iteratively applied in the model selection step to compute the model order and an initial classification for the refinement step. The refinement step uses Classification EM (CEM) to reach the final classification and the smoothing stage improves the results by means of non-linear filtering. The algorithm is applied to both simulated and real Single Look Complex data of the EMISAR mission and compared with the Wishart classification method. We use confusion matrix and kappa statistic to make the comparison for simulated data whose ground-truth is known. We apply Davies-Bouldin index to compare both classifications for real data. The results obtained for both types of data validate our algorithm and show that its performance is comparable to Wishart's in terms of classification quality.
Fil: Fernández Michelli, Juan Ignacio. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales. Universidad Nacional de La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales; Argentina
Fil: Hurtado, Martin. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales. Universidad Nacional de La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales; Argentina
Fil: Areta, Javier Alberto. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional de Río Negro; Argentina
Fil: Muravchik, Carlos Horacio. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales. Universidad Nacional de La Plata. Instituto de Investigaciones en Electrónica, Control y Procesamiento de Señales; Argentina. Universidad Nacional de la Plata. Facultad de Ingeniería. Departamento de Electrotecnia. Laboratorio de Electrónica Industrial, Control e Instrumentación; Argentina. Provincia de Buenos Aires. Gobernación. Comisión de Investigaciones Científicas; Argentina
description In this work we develop an iterative classification algorithm using complex Gaussian mixture models for the polarimetric complex SAR data. It is a non supervised algorithm which does not require training data or an initial set of classes. Additionally, it determines the model order from data, which allows representing data structure with minimum complexity. The algorithm consists of four steps: initialization, model selection, refinement and smoothing. After a simple initialization stage, the EM algorithm is iteratively applied in the model selection step to compute the model order and an initial classification for the refinement step. The refinement step uses Classification EM (CEM) to reach the final classification and the smoothing stage improves the results by means of non-linear filtering. The algorithm is applied to both simulated and real Single Look Complex data of the EMISAR mission and compared with the Wishart classification method. We use confusion matrix and kappa statistic to make the comparison for simulated data whose ground-truth is known. We apply Davies-Bouldin index to compare both classifications for real data. The results obtained for both types of data validate our algorithm and show that its performance is comparable to Wishart's in terms of classification quality.
publishDate 2016
dc.date.none.fl_str_mv 2016-07
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
http://purl.org/coar/resource_type/c_6501
info:ar-repo/semantics/articulo
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/11336/24748
Fernández Michelli, Juan Ignacio; Hurtado, Martin; Areta, Javier Alberto; Muravchik, Carlos Horacio; Unsupervised classification algorithm based on EM method for polarimetric SAR images; Elsevier Science; Isprs Journal Of Photogrammetry And Remote Sensing; 117; 7-2016; 56-65
0924-2716
CONICET Digital
CONICET
url http://hdl.handle.net/11336/24748
identifier_str_mv Fernández Michelli, Juan Ignacio; Hurtado, Martin; Areta, Javier Alberto; Muravchik, Carlos Horacio; Unsupervised classification algorithm based on EM method for polarimetric SAR images; Elsevier Science; Isprs Journal Of Photogrammetry And Remote Sensing; 117; 7-2016; 56-65
0924-2716
CONICET Digital
CONICET
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/url/http://www.sciencedirect.com/science/article/pii/S0924271616000587
info:eu-repo/semantics/altIdentifier/doi/10.1016/j.isprsjprs.2016.03.001
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
https://creativecommons.org/licenses/by-nc-nd/2.5/ar/
eu_rights_str_mv openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-nd/2.5/ar/
dc.format.none.fl_str_mv application/pdf
application/pdf
application/pdf
application/pdf
application/pdf
dc.publisher.none.fl_str_mv Elsevier Science
publisher.none.fl_str_mv Elsevier Science
dc.source.none.fl_str_mv reponame:CONICET Digital (CONICET)
instname:Consejo Nacional de Investigaciones Científicas y Técnicas
reponame_str CONICET Digital (CONICET)
collection CONICET Digital (CONICET)
instname_str Consejo Nacional de Investigaciones Científicas y Técnicas
repository.name.fl_str_mv CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicas
repository.mail.fl_str_mv dasensio@conicet.gov.ar; lcarlino@conicet.gov.ar
_version_ 1842268704462602240
score 13.13397

Publicaciones similares