|
|
Registro Completo |
Biblioteca(s): |
Embrapa Meio Ambiente. |
Data corrente: |
25/01/2016 |
Data da última atualização: |
04/01/2023 |
Tipo da produção científica: |
Artigo em Periódico Indexado |
Autoria: |
SCHULTZ, B.; IMMITZER, M.; FORMAGGIO, A. R.; SANCHES, I. D. A.; LUIZ, A. J. B.; ATZBERGER, C. |
Afiliação: |
BRUNO SCHULTZ, INPE; MARCUS IMMITZER, University of Natural Resources and Life Sciences, Viena; ANTONIO ROBERTO FORMAGGIO, INPE; IEDA DEL'ARCO SANCHES, INPE; ALFREDO JOSE BARRETO LUIZ, CNPMA; CLEMENT ATZBERGER, University of Natural Resources and Life Sciences, Viena. |
Título: |
Self-guided segmentation and classification of multi-temporal landsat 8 images for crop type mapping in southeastern Brazil. |
Ano de publicação: |
2015 |
Fonte/Imprenta: |
Remote Sensing, Basel, v. 7, n. 11, p. 14482-14508, 2015. |
ISBN: |
http://dx.doi.org/10.3390/rs71114482 |
Idioma: |
Inglês |
Conteúdo: |
Abstract: Only well-chosen segmentation parameters ensure optimum results of object-based image analysis (OBIA). Manually defining suitable parameter sets can be a time-consuming approach, not necessarily leading to optimum results; the subjectivity of the manual approach is also obvious. For this reason, in supervised segmentation as proposed by Stefanski et al. (2013) one integrates the segmentation and classification tasks. The segmentation is optimized directly with respect to the subsequent classification. In this contribution, we build on this work and developed a fully autonomous workflow for supervised object-based classification, combining image segmentation and random forest (RF) classification. Starting from a fixed set of randomly selected and manually interpreted training samples, suitable segmentation parameters are automatically identified. A sub-tropical study site located in São Paulo State (Brazil) was used to evaluate the proposed approach. Two multi-temporal Landsat 8 image mosaics were used as input (from August 2013 and January 2014) together with training samples from field visits and VHR (RapidEye) photo-interpretation. Using four test sites of 15 × 15 km2 with manually interpreted crops as independent validation samples, we demonstrate that the approach leads to robust classification results. On these samples (pixel wise, n ? 1 million) an overall accuracy (OA) of 80% could be reached while classifying five classes: sugarcane, soybean, cassava, peanut and others. We found that the overall accuracy obtained from the four test sites was only marginally lower compared to the out-of-bag OA obtained from the training samples. Amongst the five classes, sugarcane and soybean were classified best, while cassava and peanut were often misclassified due to similarity in the spatio-temporal feature space and high within-class variabilities. Interestingly, misclassified pixels were in most cases correctly identified through the RF classification margin, which is produced as a by-product to the classification map. MenosAbstract: Only well-chosen segmentation parameters ensure optimum results of object-based image analysis (OBIA). Manually defining suitable parameter sets can be a time-consuming approach, not necessarily leading to optimum results; the subjectivity of the manual approach is also obvious. For this reason, in supervised segmentation as proposed by Stefanski et al. (2013) one integrates the segmentation and classification tasks. The segmentation is optimized directly with respect to the subsequent classification. In this contribution, we build on this work and developed a fully autonomous workflow for supervised object-based classification, combining image segmentation and random forest (RF) classification. Starting from a fixed set of randomly selected and manually interpreted training samples, suitable segmentation parameters are automatically identified. A sub-tropical study site located in São Paulo State (Brazil) was used to evaluate the proposed approach. Two multi-temporal Landsat 8 image mosaics were used as input (from August 2013 and January 2014) together with training samples from field visits and VHR (RapidEye) photo-interpretation. Using four test sites of 15 × 15 km2 with manually interpreted crops as independent validation samples, we demonstrate that the approach leads to robust classification results. On these samples (pixel wise, n ? 1 million) an overall accuracy (OA) of 80% could be reached while classifying five classes: sugarcane, soybean, cassava, peanu... Mostrar Tudo |
Palavras-Chave: |
Crop mapping; Mapeamento agrícola; Multi-resolution segmentation; OBIA; OLI; Random forest; Segmentação multirresolução. |
Thesagro: |
Sensoriamento remoto. |
Thesaurus Nal: |
Brazil; Remote sensing. |
Categoria do assunto: |
X Pesquisa, Tecnologia e Engenharia |
URL: |
https://ainfo.cnptia.embrapa.br/digital/bitstream/item/137582/1/2015AP38.pdf
|
Marc: |
LEADER 02967naa a2200301 a 4500 001 2034915 005 2023-01-04 008 2015 bl uuuu u00u1 u #d 100 1 $aSCHULTZ, B. 245 $aSelf-guided segmentation and classification of multi-temporal landsat 8 images for crop type mapping in southeastern Brazil.$h[electronic resource] 260 $c2015 520 $aAbstract: Only well-chosen segmentation parameters ensure optimum results of object-based image analysis (OBIA). Manually defining suitable parameter sets can be a time-consuming approach, not necessarily leading to optimum results; the subjectivity of the manual approach is also obvious. For this reason, in supervised segmentation as proposed by Stefanski et al. (2013) one integrates the segmentation and classification tasks. The segmentation is optimized directly with respect to the subsequent classification. In this contribution, we build on this work and developed a fully autonomous workflow for supervised object-based classification, combining image segmentation and random forest (RF) classification. Starting from a fixed set of randomly selected and manually interpreted training samples, suitable segmentation parameters are automatically identified. A sub-tropical study site located in São Paulo State (Brazil) was used to evaluate the proposed approach. Two multi-temporal Landsat 8 image mosaics were used as input (from August 2013 and January 2014) together with training samples from field visits and VHR (RapidEye) photo-interpretation. Using four test sites of 15 × 15 km2 with manually interpreted crops as independent validation samples, we demonstrate that the approach leads to robust classification results. On these samples (pixel wise, n ? 1 million) an overall accuracy (OA) of 80% could be reached while classifying five classes: sugarcane, soybean, cassava, peanut and others. We found that the overall accuracy obtained from the four test sites was only marginally lower compared to the out-of-bag OA obtained from the training samples. Amongst the five classes, sugarcane and soybean were classified best, while cassava and peanut were often misclassified due to similarity in the spatio-temporal feature space and high within-class variabilities. Interestingly, misclassified pixels were in most cases correctly identified through the RF classification margin, which is produced as a by-product to the classification map. 650 $aBrazil 650 $aRemote sensing 650 $aSensoriamento remoto 653 $aCrop mapping 653 $aMapeamento agrícola 653 $aMulti-resolution segmentation 653 $aOBIA 653 $aOLI 653 $aRandom forest 653 $aSegmentação multirresolução 700 1 $aIMMITZER, M. 700 1 $aFORMAGGIO, A. R. 700 1 $aSANCHES, I. D. A. 700 1 $aLUIZ, A. J. B. 700 1 $aATZBERGER, C. 773 $tRemote Sensing, Basel$gv. 7, n. 11, p. 14482-14508, 2015.
Download
Esconder MarcMostrar Marc Completo |
Registro original: |
Embrapa Meio Ambiente (CNPMA) |
|
Biblioteca |
ID |
Origem |
Tipo/Formato |
Classificação |
Cutter |
Registro |
Volume |
Status |
URL |
Voltar
|
|
Registro Completo
Biblioteca(s): |
Embrapa Cerrados. |
Data corrente: |
29/11/2022 |
Data da última atualização: |
08/12/2022 |
Tipo da produção científica: |
Artigo em Periódico Indexado |
Circulação/Nível: |
A - 2 |
Autoria: |
ALEXANDRE, D.; NIVA, C. C.; BUSSINGER, A. P.; MARCHAO, R. L.; GATTO, A.; SILVA, R. G. da; SCHMELZ, R. M. |
Afiliação: |
DOUGLAS ALEXANDRE; CINTIA CARLA NIVA, CPAC; ANGELA P. BUSSINGER; ROBELIO LEANDRO MARCHAO, CPAC; ALCIDES GATTO; RENATA G. DA SILVA; RÜDIGER MARIA SCHMELZ. |
Título: |
First record on enchytraeids in a Savanna Tall Woodland (Cerradão) and Upper Montane Atlantic Forest in Brazil. |
Título original: |
Annals of the Brazilian Academy of Sciences |
Ano de publicação: |
2022 |
Fonte/Imprenta: |
Anais da Academia Brasileira de Ciências, v. 94, n. 4, 2022. |
Páginas: |
10 p. |
DOI: |
DOI 10.1590/0001-3765202220200892 |
Idioma: |
Inglês |
Conteúdo: |
Abstract: Brazil is considered a megadiverse country, but the soil fauna is still very poorly known. The aim of this study was to report, for the fi rst time, the abundance and genus composition of terrestrial enchytraeids (Enchytraeidae, Oligochaeta) in Savanna Tall Woodland (Cerradão) and a pasture in Cerrado Biome and in Upper Montane Atlantic Forest and a grassland in Atlantic Forest Biome. The enchytraeid density in Pasture and Cerradao was 2,036 and 18,844 (204 and 2,094, on average) individuals per square meter, respectively. At the Atlantic forest and Grassland, density was 9,666 and 12,242 individuals per square meter (1,075 and 1,471 on average). About genus composition for the studied areas, Enchytraeus and Hemienchytraeus were found in the four ecosystems evaluated, while Tupidrilus and Fridericia were found only in Cerradão and Atlantic Forest, respectively. Achaeta was absent in Upper Montane Atlantic Forest, but dominant in pasture, while Guaranidrilus was absent in Pasture, but predominant in the other ecosystems |
Palavras-Chave: |
Composição do solo; Enquitreídeo; Mata Atlântica. |
Thesagro: |
Cerrado; Minhoca; Solo. |
Categoria do assunto: |
-- |
URL: |
https://ainfo.cnptia.embrapa.br/digital/bitstream/doc/1148938/1/Cintia-Niva-first-record.pdf
|
Marc: |
LEADER 01952naa a2200301 a 4500 001 2148938 005 2022-12-08 008 2022 bl uuuu u00u1 u #d 024 7 $aDOI 10.1590/0001-3765202220200892$2DOI 100 1 $aALEXANDRE, D. 240 $aAnnals of the Brazilian Academy of Sciences 245 $aFirst record on enchytraeids in a Savanna Tall Woodland (Cerradão) and Upper Montane Atlantic Forest in Brazil.$h[electronic resource] 260 $c2022 300 $a10 p. 520 $aAbstract: Brazil is considered a megadiverse country, but the soil fauna is still very poorly known. The aim of this study was to report, for the fi rst time, the abundance and genus composition of terrestrial enchytraeids (Enchytraeidae, Oligochaeta) in Savanna Tall Woodland (Cerradão) and a pasture in Cerrado Biome and in Upper Montane Atlantic Forest and a grassland in Atlantic Forest Biome. The enchytraeid density in Pasture and Cerradao was 2,036 and 18,844 (204 and 2,094, on average) individuals per square meter, respectively. At the Atlantic forest and Grassland, density was 9,666 and 12,242 individuals per square meter (1,075 and 1,471 on average). About genus composition for the studied areas, Enchytraeus and Hemienchytraeus were found in the four ecosystems evaluated, while Tupidrilus and Fridericia were found only in Cerradão and Atlantic Forest, respectively. Achaeta was absent in Upper Montane Atlantic Forest, but dominant in pasture, while Guaranidrilus was absent in Pasture, but predominant in the other ecosystems 650 $aCerrado 650 $aMinhoca 650 $aSolo 653 $aComposição do solo 653 $aEnquitreídeo 653 $aMata Atlântica 700 1 $aNIVA, C. C. 700 1 $aBUSSINGER, A. P. 700 1 $aMARCHAO, R. L. 700 1 $aGATTO, A. 700 1 $aSILVA, R. G. da 700 1 $aSCHMELZ, R. M. 773 $tAnais da Academia Brasileira de Ciências$gv. 94, n. 4, 2022.
Download
Esconder MarcMostrar Marc Completo |
Registro original: |
Embrapa Cerrados (CPAC) |
|
Biblioteca |
ID |
Origem |
Tipo/Formato |
Classificação |
Cutter |
Registro |
Volume |
Status |
Fechar
|
Expressão de busca inválida. Verifique!!! |
|
|