Close
Metadata

%0 Conference Proceedings
%4 sid.inpe.br/mtc-m16d/2019/11.27.18.04
%2 sid.inpe.br/mtc-m16d/2019/11.27.18.04.55
%@issn 2179-4847
%A Uehara, Tatiana Dias Tardelli,
%A Correa, Sabrina Paes Leme Passos,
%A Quevedo, Renata Pacheco,
%A Körting, Thales Sehn,
%A Dutra, Luciano Vieira,
%A Rennó, Camilo Daleles,
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@electronicmailaddress tatiana.uehara@inpe.br
%@electronicmailaddress sabrina.correa@inpe.br
%@electronicmailaddress renata.quevedo@inpe.br
%@electronicmailaddress thales.korting@inpe.br
%@electronicmailaddress luciano.dutra@inpe.br
%@electronicmailaddress camilo.renno@inpe.br
%T Classification algorithms comparison for landslide scars
%B Simpósio Brasileiro de Geoinformática (GEOINFO)
%D 2019
%E Lisboa Filho, Jugurta,
%E Monteiro, Antonio Miguel Vieira,
%S Anais do 20º Simpósio Brasileiro de Geoinformática
%8 11 -13 nov. 2019
%J São José dos Campos
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%C São José dos Campos
%K geoinformatica.
%X Landslide inventory is an essential tool to support disaster risk mitigation. Using remote sensing images, it is usually obtained through pattern recognition. In this study, three classification methods are compared to detect landslides: Support Vector Machine (SVM), Artificial Neural Net (ANN) and Maximum Likelihood (ML). We used Sentinel-2A imagery, extracted and selected features for two areas in the Rolante River Catchment. The classification products showed that SVM classifier presented the best overall accuracy (OA) for Area 1 resulting in 87.143%; while for Area 2 ML showed the best OA equals to 86.831%.
%1 Instituto Nacional de Pesquisas Espaciais (INPE)
%@language pt
%3 158-169.pdf


Close