Laser Scanning Data Segmentation in Urban Areas by a Bayesian Framework
Autor(a) principal: | |
---|---|
Data de Publicação: | 2007 |
Outros Autores: | , |
Tipo de documento: | Artigo |
Idioma: | por |
Título da fonte: | Boletim de Ciências Geodésicas |
Texto Completo: | https://revistas.ufpr.br/bcg/article/view/8246 |
Resumo: | In this paper is presented a region-based methodology for Digital Elevation Modelsegmentation obtained from laser scanning data. The methodology is based on twosequential techniques, i.e., a recursive splitting technique using the quad treestructure followed by a region merging technique using the Markov Random Fieldmodel. The recursive splitting technique starts splitting the Digital Elevation Modelinto homogeneous regions. However, due to slight height differences in the DigitalElevation Model, region fragmentation can be relatively high. In order to minimizethe fragmentation, a region merging technique based on the Markov Random Fieldmodel is applied to the previously segmented data. The resulting regions are firstlystructured by using the so-called Region Adjacency Graph. Each node of theRegion Adjacency Graph represents a region of the Digital Elevation Modelsegmented and two nodes have connectivity between them if corresponding regionsshare a common boundary. Next it is assumed that the random variable related toeach node, follows the Markov Random Field model. This hypothesis allows thederivation of the posteriori probability distribution function whose solution isobtained by the Maximum a Posteriori estimation. Regions presenting highprobability of similarity are merged. Experiments carried out with laser scanningdata showed that the methodology allows to separate the objects in the DigitalElevation Model with a low amount of fragmentation. |
id |
UFPR-2_8f656834bfcf0627a13754e9d1704d90 |
---|---|
oai_identifier_str |
oai:revistas.ufpr.br:article/8246 |
network_acronym_str |
UFPR-2 |
network_name_str |
Boletim de Ciências Geodésicas |
repository_id_str |
|
spelling |
Laser Scanning Data Segmentation in Urban Areas by a Bayesian FrameworkSEGMENTAÇÃO DE DADOS DE PERFILAMENTO A LASER EM ÁREAS URBANAS UTILIZANDO UMA ABORDAGEM BAYESIANAMarkov Random Field, Modelo Digital de Elevação, Segmentação por região, QuadTree; Markov Random Field, Digital Elevation Model, Region SegmentationIn this paper is presented a region-based methodology for Digital Elevation Modelsegmentation obtained from laser scanning data. The methodology is based on twosequential techniques, i.e., a recursive splitting technique using the quad treestructure followed by a region merging technique using the Markov Random Fieldmodel. The recursive splitting technique starts splitting the Digital Elevation Modelinto homogeneous regions. However, due to slight height differences in the DigitalElevation Model, region fragmentation can be relatively high. In order to minimizethe fragmentation, a region merging technique based on the Markov Random Fieldmodel is applied to the previously segmented data. The resulting regions are firstlystructured by using the so-called Region Adjacency Graph. Each node of theRegion Adjacency Graph represents a region of the Digital Elevation Modelsegmented and two nodes have connectivity between them if corresponding regionsshare a common boundary. Next it is assumed that the random variable related toeach node, follows the Markov Random Field model. This hypothesis allows thederivation of the posteriori probability distribution function whose solution isobtained by the Maximum a Posteriori estimation. Regions presenting highprobability of similarity are merged. Experiments carried out with laser scanningdata showed that the methodology allows to separate the objects in the DigitalElevation Model with a low amount of fragmentation.Neste artigo é apresentada uma metodologia para a segmentação de um ModeloDigital de Elevação obtido a partir de um sistema de varredura a laser. Ametodologia de segmentação baseia-se na utilização das técnicas de divisãorecursiva usando a estrutura quadtree e de fusão de regiões usando o modeloMarkov Random Field. Inicialmente a técnica de divisão recursiva é usada paraparticionar o Modelo Digital de Elevação em regiões homogêneas. No entanto,devido a ligeiras diferenças de altura no Modelo Digital de Elevação, nesta etapa afragmentação das regiões pode ser relativamente alta. Para minimizar essafragmentação, uma técnica de fusão de regiões baseada no modelo Markov RandomField é aplicada nos dados segmentados. As regiões resultantes são estruturadasusando um grafo de regiões adjacentes (Region Adjacency Graph). Cada nó doRegion Adjacency Graph corresponde a uma região do Modelo Digital de Elevaçãosegmentado e dois nós tem conectividade entre eles se as duas regiõescorrespondentes compartilham de uma mesma fronteira. Em seguida, assume-se queo comportamento da variável aleatória em relação a cada nó dá se de acordo comum Markov Random Field. Esta hipótese permite a obtenção da chamadadistribuição de probabilidade a posteriori, cuja solução é obtida via estimativa Maximum a Posteriori. Regiões que apresentam alta probabilidade de fusão sãofundidas. Os experimentos realizados com os dados de perfilamento a lasermostraram que a metodologia proposta permitiu separar os objetos no ModeloDigital de Elevação com um baixo nível de fragmentação.Boletim de Ciências GeodésicasBulletin of Geodetic Sciences2007-07-03info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfhttps://revistas.ufpr.br/bcg/article/view/8246Boletim de Ciências Geodésicas; Vol 13, No 1 (2007)Bulletin of Geodetic Sciences; Vol 13, No 1 (2007)1982-21701413-4853reponame:Boletim de Ciências Geodésicasinstname:Universidade Federal do Paraná (UFPR)instacron:UFPRporhttps://revistas.ufpr.br/bcg/article/view/8246/5765Galvanin, Edinéia Aparecida dos SantosDal Poz, Aluir PorfírioPires de Souza, Aparecida Donisetiinfo:eu-repo/semantics/openAccess2007-07-03T20:30:16Zoai:revistas.ufpr.br:article/8246Revistahttps://revistas.ufpr.br/bcgPUBhttps://revistas.ufpr.br/bcg/oaiqdalmolin@ufpr.br|| danielsantos@ufpr.br||qdalmolin@ufpr.br|| danielsantos@ufpr.br1982-21701413-4853opendoar:2007-07-03T20:30:16Boletim de Ciências Geodésicas - Universidade Federal do Paraná (UFPR)false |
dc.title.none.fl_str_mv |
Laser Scanning Data Segmentation in Urban Areas by a Bayesian Framework SEGMENTAÇÃO DE DADOS DE PERFILAMENTO A LASER EM ÁREAS URBANAS UTILIZANDO UMA ABORDAGEM BAYESIANA |
title |
Laser Scanning Data Segmentation in Urban Areas by a Bayesian Framework |
spellingShingle |
Laser Scanning Data Segmentation in Urban Areas by a Bayesian Framework Galvanin, Edinéia Aparecida dos Santos Markov Random Field, Modelo Digital de Elevação, Segmentação por região, QuadTree; Markov Random Field, Digital Elevation Model, Region Segmentation |
title_short |
Laser Scanning Data Segmentation in Urban Areas by a Bayesian Framework |
title_full |
Laser Scanning Data Segmentation in Urban Areas by a Bayesian Framework |
title_fullStr |
Laser Scanning Data Segmentation in Urban Areas by a Bayesian Framework |
title_full_unstemmed |
Laser Scanning Data Segmentation in Urban Areas by a Bayesian Framework |
title_sort |
Laser Scanning Data Segmentation in Urban Areas by a Bayesian Framework |
author |
Galvanin, Edinéia Aparecida dos Santos |
author_facet |
Galvanin, Edinéia Aparecida dos Santos Dal Poz, Aluir Porfírio Pires de Souza, Aparecida Doniseti |
author_role |
author |
author2 |
Dal Poz, Aluir Porfírio Pires de Souza, Aparecida Doniseti |
author2_role |
author author |
dc.contributor.author.fl_str_mv |
Galvanin, Edinéia Aparecida dos Santos Dal Poz, Aluir Porfírio Pires de Souza, Aparecida Doniseti |
dc.subject.por.fl_str_mv |
Markov Random Field, Modelo Digital de Elevação, Segmentação por região, QuadTree; Markov Random Field, Digital Elevation Model, Region Segmentation |
topic |
Markov Random Field, Modelo Digital de Elevação, Segmentação por região, QuadTree; Markov Random Field, Digital Elevation Model, Region Segmentation |
description |
In this paper is presented a region-based methodology for Digital Elevation Modelsegmentation obtained from laser scanning data. The methodology is based on twosequential techniques, i.e., a recursive splitting technique using the quad treestructure followed by a region merging technique using the Markov Random Fieldmodel. The recursive splitting technique starts splitting the Digital Elevation Modelinto homogeneous regions. However, due to slight height differences in the DigitalElevation Model, region fragmentation can be relatively high. In order to minimizethe fragmentation, a region merging technique based on the Markov Random Fieldmodel is applied to the previously segmented data. The resulting regions are firstlystructured by using the so-called Region Adjacency Graph. Each node of theRegion Adjacency Graph represents a region of the Digital Elevation Modelsegmented and two nodes have connectivity between them if corresponding regionsshare a common boundary. Next it is assumed that the random variable related toeach node, follows the Markov Random Field model. This hypothesis allows thederivation of the posteriori probability distribution function whose solution isobtained by the Maximum a Posteriori estimation. Regions presenting highprobability of similarity are merged. Experiments carried out with laser scanningdata showed that the methodology allows to separate the objects in the DigitalElevation Model with a low amount of fragmentation. |
publishDate |
2007 |
dc.date.none.fl_str_mv |
2007-07-03 |
dc.type.none.fl_str_mv |
|
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
https://revistas.ufpr.br/bcg/article/view/8246 |
url |
https://revistas.ufpr.br/bcg/article/view/8246 |
dc.language.iso.fl_str_mv |
por |
language |
por |
dc.relation.none.fl_str_mv |
https://revistas.ufpr.br/bcg/article/view/8246/5765 |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.publisher.none.fl_str_mv |
Boletim de Ciências Geodésicas Bulletin of Geodetic Sciences |
publisher.none.fl_str_mv |
Boletim de Ciências Geodésicas Bulletin of Geodetic Sciences |
dc.source.none.fl_str_mv |
Boletim de Ciências Geodésicas; Vol 13, No 1 (2007) Bulletin of Geodetic Sciences; Vol 13, No 1 (2007) 1982-2170 1413-4853 reponame:Boletim de Ciências Geodésicas instname:Universidade Federal do Paraná (UFPR) instacron:UFPR |
instname_str |
Universidade Federal do Paraná (UFPR) |
instacron_str |
UFPR |
institution |
UFPR |
reponame_str |
Boletim de Ciências Geodésicas |
collection |
Boletim de Ciências Geodésicas |
repository.name.fl_str_mv |
Boletim de Ciências Geodésicas - Universidade Federal do Paraná (UFPR) |
repository.mail.fl_str_mv |
qdalmolin@ufpr.br|| danielsantos@ufpr.br||qdalmolin@ufpr.br|| danielsantos@ufpr.br |
_version_ |
1799771721217081344 |