<?xml version="1.0" encoding="ISO-8859-1"?>
<metadatalist>
	<metadata ReferenceType="Journal Article">
		<site>plutao.sid.inpe.br 800</site>
		<holdercode>{isadg {BR SPINPE} ibi 8JMKD3MGPCW/3DT298S}</holdercode>
		<identifier>8JMKD3MGP3W/45U858J</identifier>
		<repository>sid.inpe.br/plutao/2021/12.09.14.40.28</repository>
		<lastupdate>2021:12.14.18.07.10 dpi.inpe.br/plutao@80/2008/08.19.15.01 lattes</lastupdate>
		<metadatarepository>sid.inpe.br/plutao/2021/12.09.14.40.29</metadatarepository>
		<metadatalastupdate>2022:04.03.22.27.14 sid.inpe.br/bibdigital@80/2006/04.07.15.50 administrator</metadatalastupdate>
		<doi>10.1117/1.JRS.15.044504</doi>
		<issn>1931-3195</issn>
		<label>lattes: 5123287769635741 3 NevesKöFoSoGiHe:2021:HiMaBr</label>
		<citationkey>NevesKöFoSoGiHe:2021:HiMaBr</citationkey>
		<title>Hierarchical mapping of Brazilian Savanna (Cerrado) physiognomies based on deep learning</title>
		<year>2021</year>
		<typeofwork>journal article</typeofwork>
		<secondarytype>PRE PI</secondarytype>
		<numberoffiles>1</numberoffiles>
		<size>8895 KiB</size>
		<author>Neves, Alana K.,</author>
		<author>Körting, Thales Sehn,</author>
		<author>Fonseca, Leila Maria Garcia,</author>
		<author>Soares, Anderson Reis,</author>
		<author>Girolamo Neto, Cesare Di,</author>
		<author>Heipke, Christian,</author>
		<resumeid></resumeid>
		<resumeid></resumeid>
		<resumeid>8JMKD3MGP5W/3C9JHLD</resumeid>
		<orcid>0000-0002-0301-944X</orcid>
		<orcid>0000-0002-0876-0501</orcid>
		<orcid>0000-0001-6057-7387</orcid>
		<orcid>0000-0001-6513-2192</orcid>
		<orcid>0000-0003-2533-4640</orcid>
		<orcid>0000-0002-7007-9549</orcid>
		<group>SER-SRE-DIPGR-INPE-MCTI-GOV-BR</group>
		<group>DIOTG-CGCT-INPE-MCTI-GOV-BR</group>
		<group>DIOTG-CGCT-INPE-MCTI-GOV-BR</group>
		<group>DIOTG-CGCT-INPE-MCTI-GOV-BR</group>
		<group>DIOTG-CGCT-INPE-MCTI-GOV-BR</group>
		<affiliation>Instituto Nacional de Pesquisas Espaciais (INPE)</affiliation>
		<affiliation>Instituto Nacional de Pesquisas Espaciais (INPE)</affiliation>
		<affiliation>Instituto Nacional de Pesquisas Espaciais (INPE)</affiliation>
		<affiliation>Instituto Nacional de Pesquisas Espaciais (INPE)</affiliation>
		<affiliation>Instituto Nacional de Pesquisas Espaciais (INPE)</affiliation>
		<affiliation>Leibniz Universität Hannover</affiliation>
		<electronicmailaddress>alana.kasahara@gmail.com</electronicmailaddress>
		<electronicmailaddress>thales.korting.@inpe.br</electronicmailaddress>
		<electronicmailaddress>leila.fonseca@inpe.br</electronicmailaddress>
		<electronicmailaddress>anderson.soares@inpe.br</electronicmailaddress>
		<electronicmailaddress>cesare.neto@gmail.com</electronicmailaddress>
		<journal>Journal of Applied Remote Sensing</journal>
		<volume>15</volume>
		<number>04</number>
		<pages>e044504</pages>
		<secondarymark>A2_GEOGRAFIA B1_GEOCIÊNCIAS B1_CIÊNCIAS_AGRÁRIAS_I B1_BIODIVERSIDADE B5_CIÊNCIAS_AMBIENTAIS</secondarymark>
		<transferableflag>1</transferableflag>
		<contenttype>External Contribution</contenttype>
		<versiontype>publisher</versiontype>
		<keywords>Savanna, Cerrado, physiognomy, semantic segmentation, spectral channels, protected area.</keywords>
		<abstract>The Brazilian Savanna, also known as Cerrado, is considered a global hotspot for biodiversity conservation. The detailed mapping of vegetation types, called physiognomies, is still a challenge due to their high spectral similarity and spatial variability. There are three major ecosystem groups (forest, savanna, and grassland), which can be hierarchically subdivided into 25 detailed physiognomies, according to a well-known classification system. We used an adapted U-net architecture to process a WorldView-2 image with 2-m spatial resolution to hierarchically classify the physiognomies of a Cerrado protected area based on deep learning techniques. Several spectral channels were tested as input datasets to classify the three major ecosystem groups (first level of classification). The dataset composed of RGB bands plus 2-band enhanced vegetation index (EVI2) achieved the best performance and was used to perform the hierarchical classification. In the first level of classification, the overall accuracy was 92.8%. On the other hand, for the savanna and grassland detailed physiognomies (second level of classification), 86.1% and 85.0% were reached, respectively. As the first work that intended to classify Cerrado physiognomies in this level of detail using deep learning, our accuracy rates outperformed others that applied traditional machine learning algorithms for this task. © The Authors. Published by SPIE under a Creative Commons Attribution 4.0 International License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI. [DOI: 10.1117/1.JRS.15.044504].</abstract>
		<area>SRE</area>
		<language>pt</language>
		<targetfile>neves_hierarchical.pdf</targetfile>
		<readergroup>administrator</readergroup>
		<readergroup>lattes</readergroup>
		<visibility>shown</visibility>
		<archivingpolicy>allowpublisher allowfinaldraft</archivingpolicy>
		<documentstage>not transferred</documentstage>
		<nexthigherunit>8JMKD3MGPCW/3F3NU5S</nexthigherunit>
		<nexthigherunit>8JMKD3MGPCW/46KUATE</nexthigherunit>
		<citingitemlist>sid.inpe.br/bibdigital/2013/10.18.22.34 2</citingitemlist>
		<url>https://caps.luminad.com:8443/stockage/stock/SPIE/LDL-SPIE-JARS-210442/JARS-210442_online.pdf</url>
		<dissemination>WEBSCI; PORTALCAPES; COMPENDEX; SCOPUS.</dissemination>
		<hostcollection>dpi.inpe.br/plutao@80/2008/08.19.15.01</hostcollection>
		<username>simone</username>
		<lasthostcollection>dpi.inpe.br/plutao@80/2008/08.19.15.01</lasthostcollection>
		<url>http://plutao.sid.inpe.br/rep-/sid.inpe.br/plutao/2021/12.09.14.40.28</url>
	</metadata>
</metadatalist>