A new image segmentation framework based on two-dimensional hidden Markov models

Autores
Baumgartner, Josef Sylvester; Flesia, Ana Georgina; Gimenez Romero, Javier Alejandro; Pucheta, Julián Antonio
Año de publicación
2015
Idioma
inglés
Tipo de recurso
artículo
Estado
versión publicada
Descripción
Image segmentation is one of the fundamental problems in computer vision. In this work, we present a new segmentation algorithm that is based on the theory of two-dimensional hidden Markov models (2D-HMM). Unlike most 2D-HMM approaches we do not apply the Viterbi algorithm; instead we present a computationally efficient algorithm that propagates the state probabilities through the image. Our algorithm, called Complete Enumeration Iteration (CEP), is flexible in the sense that it allows the use of different probability distributions as emibion probabilities. Not only do we compare the performance of different probability functions plugged into our framework but also propose three methods to update the distributions of each state "online" during the segmentation proceb. We compare our algorithm with a 2D-HMM standard algorithm and Iterated Conditional Modes (ICM) using real world images like a radiography or a satellite image as well as synthetic images. The experimental results are evaluated by the kappa coefficient (κ). In those cases where the average κ coefficient is higher than 0.7 we observe an average relative improvement of 8% of CEP with respect to the benchmark algorithms. For all other segmentation tasks CEP shows no significant improvement. Besides that, we demonstrate how the choice of the emibion probability can have great influence on the segmentation results. Surprisingly, we observe that the normal distribution is an appropriate density function for many segmentation tasks.
Fil: Baumgartner, Josef Sylvester. Universidad Nacional de Córdoba. Facultad de Cs.exactas Físicas y Naturales. Departamento de Electronica; Argentina. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina
Fil: Flesia, Ana Georgina. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Córdoba. Centro de Investigación y Estudios de Matemática. Universidad Nacional de Córdoba. Centro de Investigación y Estudios de Matemática; Argentina
Fil: Gimenez Romero, Javier Alejandro. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - San Juan. Instituto de Automática. Universidad Nacional de San Juan. Facultad de Ingeniería. Instituto de Automática; Argentina
Fil: Pucheta, Julián Antonio. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional de Córdoba. Facultad de Cs.exactas Físicas y Naturales. Departamento de Electronica; Argentina
Materia
Hidden Markov Models
Image Segmentation
Kappa Coefficient
Probability Density Function
Viterbi Training
Nivel de accesibilidad
acceso abierto
Condiciones de uso
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
Repositorio
CONICET Digital (CONICET)
Institución
Consejo Nacional de Investigaciones Científicas y Técnicas
OAI Identificador
oai:ri.conicet.gov.ar:11336/58360
Acceder
id CONICETDig_37fb13d4a00b8102407a2eac6989324f
oai_identifier_str oai:ri.conicet.gov.ar:11336/58360
network_acronym_str CONICETDig
repository_id_str 3498
network_name_str CONICET Digital (CONICET)
spelling A new image segmentation framework based on two-dimensional hidden Markov modelsBaumgartner, Josef SylvesterFlesia, Ana GeorginaGimenez Romero, Javier AlejandroPucheta, Julián AntonioHidden Markov ModelsImage SegmentationKappa CoefficientProbability Density FunctionViterbi Traininghttps://purl.org/becyt/ford/1.1https://purl.org/becyt/ford/1Image segmentation is one of the fundamental problems in computer vision. In this work, we present a new segmentation algorithm that is based on the theory of two-dimensional hidden Markov models (2D-HMM). Unlike most 2D-HMM approaches we do not apply the Viterbi algorithm; instead we present a computationally efficient algorithm that propagates the state probabilities through the image. Our algorithm, called Complete Enumeration Iteration (CEP), is flexible in the sense that it allows the use of different probability distributions as emibion probabilities. Not only do we compare the performance of different probability functions plugged into our framework but also propose three methods to update the distributions of each state "online" during the segmentation proceb. We compare our algorithm with a 2D-HMM standard algorithm and Iterated Conditional Modes (ICM) using real world images like a radiography or a satellite image as well as synthetic images. The experimental results are evaluated by the kappa coefficient (κ). In those cases where the average κ coefficient is higher than 0.7 we observe an average relative improvement of 8% of CEP with respect to the benchmark algorithms. For all other segmentation tasks CEP shows no significant improvement. Besides that, we demonstrate how the choice of the emibion probability can have great influence on the segmentation results. Surprisingly, we observe that the normal distribution is an appropriate density function for many segmentation tasks.Fil: Baumgartner, Josef Sylvester. Universidad Nacional de Córdoba. Facultad de Cs.exactas Físicas y Naturales. Departamento de Electronica; Argentina. Consejo Nacional de Investigaciones Científicas y Técnicas; ArgentinaFil: Flesia, Ana Georgina. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Córdoba. Centro de Investigación y Estudios de Matemática. Universidad Nacional de Córdoba. Centro de Investigación y Estudios de Matemática; ArgentinaFil: Gimenez Romero, Javier Alejandro. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - San Juan. Instituto de Automática. Universidad Nacional de San Juan. Facultad de Ingeniería. Instituto de Automática; ArgentinaFil: Pucheta, Julián Antonio. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional de Córdoba. Facultad de Cs.exactas Físicas y Naturales. Departamento de Electronica; ArgentinaIOS Press2015-12info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionhttp://purl.org/coar/resource_type/c_6501info:ar-repo/semantics/articuloapplication/pdfapplication/pdfhttp://hdl.handle.net/11336/58360Baumgartner, Josef Sylvester; Flesia, Ana Georgina; Gimenez Romero, Javier Alejandro; Pucheta, Julián Antonio; A new image segmentation framework based on two-dimensional hidden Markov models; IOS Press; Integrated Computer-aided Engineering; 23; 1; 12-2015; 1-131069-2509CONICET DigitalCONICETenginfo:eu-repo/semantics/altIdentifier/url/https://content.iospress.com/articles/integrated-computer-aided-engineering/ica497info:eu-repo/semantics/altIdentifier/doi/10.3233/ICA-150497info:eu-repo/semantics/openAccesshttps://creativecommons.org/licenses/by-nc-sa/2.5/ar/reponame:CONICET Digital (CONICET)instname:Consejo Nacional de Investigaciones Científicas y Técnicas2025-09-03T09:47:35Zoai:ri.conicet.gov.ar:11336/58360instacron: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 09:47:35.992CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicasfalse
dc.title.none.fl_str_mv A new image segmentation framework based on two-dimensional hidden Markov models
title A new image segmentation framework based on two-dimensional hidden Markov models
spellingShingle A new image segmentation framework based on two-dimensional hidden Markov models
Baumgartner, Josef Sylvester
Hidden Markov Models
Image Segmentation
Kappa Coefficient
Probability Density Function
Viterbi Training
title_short A new image segmentation framework based on two-dimensional hidden Markov models
title_full A new image segmentation framework based on two-dimensional hidden Markov models
title_fullStr A new image segmentation framework based on two-dimensional hidden Markov models
title_full_unstemmed A new image segmentation framework based on two-dimensional hidden Markov models
title_sort A new image segmentation framework based on two-dimensional hidden Markov models
dc.creator.none.fl_str_mv Baumgartner, Josef Sylvester
Flesia, Ana Georgina
Gimenez Romero, Javier Alejandro
Pucheta, Julián Antonio
author Baumgartner, Josef Sylvester
author_facet Baumgartner, Josef Sylvester
Flesia, Ana Georgina
Gimenez Romero, Javier Alejandro
Pucheta, Julián Antonio
author_role author
author2 Flesia, Ana Georgina
Gimenez Romero, Javier Alejandro
Pucheta, Julián Antonio
author2_role author
author
author
dc.subject.none.fl_str_mv Hidden Markov Models
Image Segmentation
Kappa Coefficient
Probability Density Function
Viterbi Training
topic Hidden Markov Models
Image Segmentation
Kappa Coefficient
Probability Density Function
Viterbi Training
purl_subject.fl_str_mv https://purl.org/becyt/ford/1.1
https://purl.org/becyt/ford/1
dc.description.none.fl_txt_mv Image segmentation is one of the fundamental problems in computer vision. In this work, we present a new segmentation algorithm that is based on the theory of two-dimensional hidden Markov models (2D-HMM). Unlike most 2D-HMM approaches we do not apply the Viterbi algorithm; instead we present a computationally efficient algorithm that propagates the state probabilities through the image. Our algorithm, called Complete Enumeration Iteration (CEP), is flexible in the sense that it allows the use of different probability distributions as emibion probabilities. Not only do we compare the performance of different probability functions plugged into our framework but also propose three methods to update the distributions of each state "online" during the segmentation proceb. We compare our algorithm with a 2D-HMM standard algorithm and Iterated Conditional Modes (ICM) using real world images like a radiography or a satellite image as well as synthetic images. The experimental results are evaluated by the kappa coefficient (κ). In those cases where the average κ coefficient is higher than 0.7 we observe an average relative improvement of 8% of CEP with respect to the benchmark algorithms. For all other segmentation tasks CEP shows no significant improvement. Besides that, we demonstrate how the choice of the emibion probability can have great influence on the segmentation results. Surprisingly, we observe that the normal distribution is an appropriate density function for many segmentation tasks.
Fil: Baumgartner, Josef Sylvester. Universidad Nacional de Córdoba. Facultad de Cs.exactas Físicas y Naturales. Departamento de Electronica; Argentina. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina
Fil: Flesia, Ana Georgina. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Córdoba. Centro de Investigación y Estudios de Matemática. Universidad Nacional de Córdoba. Centro de Investigación y Estudios de Matemática; Argentina
Fil: Gimenez Romero, Javier Alejandro. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - San Juan. Instituto de Automática. Universidad Nacional de San Juan. Facultad de Ingeniería. Instituto de Automática; Argentina
Fil: Pucheta, Julián Antonio. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Universidad Nacional de Córdoba. Facultad de Cs.exactas Físicas y Naturales. Departamento de Electronica; Argentina
description Image segmentation is one of the fundamental problems in computer vision. In this work, we present a new segmentation algorithm that is based on the theory of two-dimensional hidden Markov models (2D-HMM). Unlike most 2D-HMM approaches we do not apply the Viterbi algorithm; instead we present a computationally efficient algorithm that propagates the state probabilities through the image. Our algorithm, called Complete Enumeration Iteration (CEP), is flexible in the sense that it allows the use of different probability distributions as emibion probabilities. Not only do we compare the performance of different probability functions plugged into our framework but also propose three methods to update the distributions of each state "online" during the segmentation proceb. We compare our algorithm with a 2D-HMM standard algorithm and Iterated Conditional Modes (ICM) using real world images like a radiography or a satellite image as well as synthetic images. The experimental results are evaluated by the kappa coefficient (κ). In those cases where the average κ coefficient is higher than 0.7 we observe an average relative improvement of 8% of CEP with respect to the benchmark algorithms. For all other segmentation tasks CEP shows no significant improvement. Besides that, we demonstrate how the choice of the emibion probability can have great influence on the segmentation results. Surprisingly, we observe that the normal distribution is an appropriate density function for many segmentation tasks.
publishDate 2015
dc.date.none.fl_str_mv 2015-12
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/11336/58360
Baumgartner, Josef Sylvester; Flesia, Ana Georgina; Gimenez Romero, Javier Alejandro; Pucheta, Julián Antonio; A new image segmentation framework based on two-dimensional hidden Markov models; IOS Press; Integrated Computer-aided Engineering; 23; 1; 12-2015; 1-13
1069-2509
CONICET Digital
CONICET
url http://hdl.handle.net/11336/58360
identifier_str_mv Baumgartner, Josef Sylvester; Flesia, Ana Georgina; Gimenez Romero, Javier Alejandro; Pucheta, Julián Antonio; A new image segmentation framework based on two-dimensional hidden Markov models; IOS Press; Integrated Computer-aided Engineering; 23; 1; 12-2015; 1-13
1069-2509
CONICET Digital
CONICET
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/url/https://content.iospress.com/articles/integrated-computer-aided-engineering/ica497
info:eu-repo/semantics/altIdentifier/doi/10.3233/ICA-150497
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
eu_rights_str_mv openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv IOS Press
publisher.none.fl_str_mv IOS Press
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_ 1842268869961449472
score 13.13397

Publicaciones similares