Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests
Autor(a) principal: | |
---|---|
Data de Publicação: | 2004 |
Outros Autores: | , , |
Tipo de documento: | Artigo |
Idioma: | eng |
Título da fonte: | Repositório Institucional da UFLA |
Texto Completo: | http://www.sciencedirect.com/science/article/pii/S0303243404000194 http://repositorio.ufla.br/jspui/handle/1/856 |
Resumo: | This paper presents a case study on the use of features derived from remote sensing data for mapping the highly fragmented semideciduous Atlantic forest in Brazil. Innovative aspects of this research include the evaluation of different feature sets in order to improve land covermapping. The feature sets were defined based on expert knowledge and on data mining techniques to be input to traditional and machine learning algorithms for pattern recognition, viz. maximum likelihood, univariate decision trees, multivariate decision trees, and neural networks. The results showed that the maximum likelihood classification using temporal texture descriptors as extracted with wavelet transforms was most accurate to classify the semideciduous Atlantic forest. In this study, a special accuracy measure was used: the so-called class mapping accuracy. Maximum likelihood performed relatively well, with forest mapping accuracies ranging from 34.5 to 51.3%. In contrast, accuracies for neural networks ranged from 19.0 to 45.2%. Classification confusion occurred mainly with coffee and eucalyptus plantations. Univariate trees provided the most robust results for different feature sets, with accuracies ranging from 39.6 to 46.7%. Temporal information of vegetation indices was more important than image texture, terrain topography and raw spectral information for discriminating semideciduous Atlantic forest. |
id |
UFLA_8c43226832efd62b1bf760094c3823ee |
---|---|
oai_identifier_str |
oai:localhost:1/856 |
network_acronym_str |
UFLA |
network_name_str |
Repositório Institucional da UFLA |
repository_id_str |
|
spelling |
Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forestsForest classificationFeature setsClassifiersArtificial intelligenceThis paper presents a case study on the use of features derived from remote sensing data for mapping the highly fragmented semideciduous Atlantic forest in Brazil. Innovative aspects of this research include the evaluation of different feature sets in order to improve land covermapping. The feature sets were defined based on expert knowledge and on data mining techniques to be input to traditional and machine learning algorithms for pattern recognition, viz. maximum likelihood, univariate decision trees, multivariate decision trees, and neural networks. The results showed that the maximum likelihood classification using temporal texture descriptors as extracted with wavelet transforms was most accurate to classify the semideciduous Atlantic forest. In this study, a special accuracy measure was used: the so-called class mapping accuracy. Maximum likelihood performed relatively well, with forest mapping accuracies ranging from 34.5 to 51.3%. In contrast, accuracies for neural networks ranged from 19.0 to 45.2%. Classification confusion occurred mainly with coffee and eucalyptus plantations. Univariate trees provided the most robust results for different feature sets, with accuracies ranging from 39.6 to 46.7%. Temporal information of vegetation indices was more important than image texture, terrain topography and raw spectral information for discriminating semideciduous Atlantic forest.Elsevier2013-08-05T20:24:08Z2013-08-05T20:24:08Z2004info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articleCARVALHO, L. M. T. et al. Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests. International Journal of Applied Earth Observation and Geoinformation, Enschede, v. 5, n. 3, p. 173-186, Sept. 2004.http://www.sciencedirect.com/science/article/pii/S0303243404000194http://repositorio.ufla.br/jspui/handle/1/856International Journal of Applied Earth Observation and Geoinformationreponame:Repositório Institucional da UFLAinstname:Universidade Federal de Lavras (UFLA)instacron:UFLACarvalho, Luis Marcelo Tavares deClevers, Jan G.P.W.Skidmore, Andrew K.Jong, Steven M. deinfo:eu-repo/semantics/openAccesseng2013-09-04T19:18:57Zoai:localhost:1/856Repositório InstitucionalPUBhttp://repositorio.ufla.br/oai/requestnivaldo@ufla.br || repositorio.biblioteca@ufla.bropendoar:2013-09-04T19:18:57Repositório Institucional da UFLA - Universidade Federal de Lavras (UFLA)false |
dc.title.none.fl_str_mv |
Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests |
title |
Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests |
spellingShingle |
Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests Carvalho, Luis Marcelo Tavares de Forest classification Feature sets Classifiers Artificial intelligence |
title_short |
Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests |
title_full |
Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests |
title_fullStr |
Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests |
title_full_unstemmed |
Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests |
title_sort |
Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests |
author |
Carvalho, Luis Marcelo Tavares de |
author_facet |
Carvalho, Luis Marcelo Tavares de Clevers, Jan G.P.W. Skidmore, Andrew K. Jong, Steven M. de |
author_role |
author |
author2 |
Clevers, Jan G.P.W. Skidmore, Andrew K. Jong, Steven M. de |
author2_role |
author author author |
dc.contributor.author.fl_str_mv |
Carvalho, Luis Marcelo Tavares de Clevers, Jan G.P.W. Skidmore, Andrew K. Jong, Steven M. de |
dc.subject.por.fl_str_mv |
Forest classification Feature sets Classifiers Artificial intelligence |
topic |
Forest classification Feature sets Classifiers Artificial intelligence |
description |
This paper presents a case study on the use of features derived from remote sensing data for mapping the highly fragmented semideciduous Atlantic forest in Brazil. Innovative aspects of this research include the evaluation of different feature sets in order to improve land covermapping. The feature sets were defined based on expert knowledge and on data mining techniques to be input to traditional and machine learning algorithms for pattern recognition, viz. maximum likelihood, univariate decision trees, multivariate decision trees, and neural networks. The results showed that the maximum likelihood classification using temporal texture descriptors as extracted with wavelet transforms was most accurate to classify the semideciduous Atlantic forest. In this study, a special accuracy measure was used: the so-called class mapping accuracy. Maximum likelihood performed relatively well, with forest mapping accuracies ranging from 34.5 to 51.3%. In contrast, accuracies for neural networks ranged from 19.0 to 45.2%. Classification confusion occurred mainly with coffee and eucalyptus plantations. Univariate trees provided the most robust results for different feature sets, with accuracies ranging from 39.6 to 46.7%. Temporal information of vegetation indices was more important than image texture, terrain topography and raw spectral information for discriminating semideciduous Atlantic forest. |
publishDate |
2004 |
dc.date.none.fl_str_mv |
2004 2013-08-05T20:24:08Z 2013-08-05T20:24:08Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/article |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
CARVALHO, L. M. T. et al. Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests. International Journal of Applied Earth Observation and Geoinformation, Enschede, v. 5, n. 3, p. 173-186, Sept. 2004. http://www.sciencedirect.com/science/article/pii/S0303243404000194 http://repositorio.ufla.br/jspui/handle/1/856 |
identifier_str_mv |
CARVALHO, L. M. T. et al. Selection of imagery data and classifiers for mapping Brazilian semideciduous Atlantic forests. International Journal of Applied Earth Observation and Geoinformation, Enschede, v. 5, n. 3, p. 173-186, Sept. 2004. |
url |
http://www.sciencedirect.com/science/article/pii/S0303243404000194 http://repositorio.ufla.br/jspui/handle/1/856 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.publisher.none.fl_str_mv |
Elsevier |
publisher.none.fl_str_mv |
Elsevier |
dc.source.none.fl_str_mv |
International Journal of Applied Earth Observation and Geoinformation reponame:Repositório Institucional da UFLA instname:Universidade Federal de Lavras (UFLA) instacron:UFLA |
instname_str |
Universidade Federal de Lavras (UFLA) |
instacron_str |
UFLA |
institution |
UFLA |
reponame_str |
Repositório Institucional da UFLA |
collection |
Repositório Institucional da UFLA |
repository.name.fl_str_mv |
Repositório Institucional da UFLA - Universidade Federal de Lavras (UFLA) |
repository.mail.fl_str_mv |
nivaldo@ufla.br || repositorio.biblioteca@ufla.br |
_version_ |
1807835145980346368 |