Close

1. Identity statement
Reference TypeConference Paper (Conference Proceedings)
Sitemarte.sid.inpe.br
Holder Codeisadg {BR SPINPE} ibi 8JMKD3MGPCW/3DT298S
Repositoryltid.inpe.br/sbsr/2004/11.20.21.24
Last Update2005:03.25.13.18.08 (UTC) administrator
Metadata Repositoryltid.inpe.br/sbsr/2004/11.20.21.24.40
Metadata Last Update2018:06.06.02.42.38 (UTC) administrator
Secondary KeyINPE-12766-PRE/8056
ISBN85-17-00018-8
Citation KeyOliveiraDutrRenn:2005:ApMéEx
TitleAplicação de métodos de extração e seleção de atributos para classificação de regiões
FormatCD-ROM, On-line.
Year2005
Access Date2024, May 19
Secondary TypePRE CN
Number of Files1
Size395 KiB
2. Context
Author1 Oliveira, Joanito de Andrade
2 Dutra, Luciano Vieira
3 Rennó, Camilo Daleles
Resume Identifier1
2 8JMKD3MGP5W/3C9JHMA
3 8JMKD3MGP5W/3C9JGN2
Group1 DPI-INPE-MCT-BR
2 DPI-INPE-MCT-BR
3 DPI-INPE-MCT-BR
Affiliation1 Instituto Nacional de Pesquisas Espaciais (INPE)
2 Instituto Nacional de Pesquisas Espaciais (INPE)
3 Instituto Nacional de Pesquisas Espaciais (INPE)
Author e-Mail Address1 joanito@dpi.inpe.br
2 dutra@dpi.inpe.br
3 camilo@dpi.inpe.br
EditorEpiphanio, José Carlos Neves
Fonseca, Leila Maria Garcia
e-Mail Addressjoanito@dpi.inpe.br
Conference NameSimpósio Brasileiro de Sensoriamento Remoto, 12 (SBSR)
Conference LocationGoiânia
Date16-21 abr. 2005
PublisherINPE
Publisher CitySão José dos Campos
Pages4201-4208
Book TitleAnais
Tertiary TypeArtigos
OrganizationInstituto Nacional de Pesquisas Espaciais
History (UTC)2005-06-24 18:00:54 :: sbsr -> administrator ::
2009-06-03 16:01:32 :: administrator -> sbsr ::
2009-06-30 14:26:02 :: sbsr -> erich@sid.inpe.br ::
2010-05-14 02:46:39 :: erich@sid.inpe.br -> marciana ::
2011-02-16 13:59:59 :: marciana -> administrator :: 2005
2018-06-06 02:42:38 :: administrator -> :: 2005
3. Content and structure
Is the master or a copy?is the master
Content Stagecompleted
Transferable1
Keywordsfeature extraction
feature selection
region classification
AbstractGood feature extraction and selection methods are crucial for an adequate performance of remote sensing data classification methods. When data dimensionality is very high, a search strategy should be used to select the set of features that gives the minimum classification error or minimize an alternate cheaper criterion related to the classification error. In this paper different search strategies are compared for the case where a region classifier is used. Landsat imagery is first segmented using standard region growing technique. Afterwards a large number of features are extracted from regions, and a smaller set of features is selected by means of three types of search strategy. The set of features selected by each criterion is classified based on minimum Mahalanobis distance and classification accuracy is calculated. As there is no deterministic relation between standard feature selection methods and classification error, is possible to conclude that all search strategies should be used to narrow the number of choices assessments and taking the final decision based on classification error.
AreaSRE
TypeProcessamento de Imagens
doc Directory Contentaccess
source Directory Content
XIISBSR_joanito.doc 04/12/2004 15:30 640.0 KiB 
agreement Directory Contentthere are no files
4. Conditions of access and use
data URLhttp://urlib.net/ibi/ltid.inpe.br/sbsr/2004/11.20.21.24
zipped data URLhttp://urlib.net/zip/ltid.inpe.br/sbsr/2004/11.20.21.24
Languagept
Target File4201.pdf
User Groupadministrator
erich@sid.inpe.br
Visibilityshown
5. Allied materials
Mirror Repositorydpi.inpe.br/marte@80/2007/10.17.19.59
Next Higher Units8JMKD3MGPCW/3EQCCU5
Citing Item Listsid.inpe.br/mtc-m21/2012/07.13.14.42 2
sid.inpe.br/mtc-m21/2012/07.13.14.53.50 1
Host Collectiondpi.inpe.br/banon/2003/12.10.19.30
6. Notes
Mark1
Empty Fieldsarchivingpolicy archivist callnumber contenttype copyholder copyright creatorhistory descriptionlevel dissemination documentstage doi edition identifier issn label lineage nextedition notes numberofvolumes orcid parameterlist parentrepositories previousedition previouslowerunit progress project readergroup readpermission rightsholder schedulinginformation secondarydate secondarymark serieseditor session shorttitle sponsor subject tertiarymark url versiontype volume


Close