Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.

Detalhes bibliográficos
Autor(a) principal: FERREIRA, M. P.
Data de Publicação: 2020
Outros Autores: ALMEIDA, D. R. A. de, PAPA, D. de A., MINERVINO, J. B. S., VERAS, H. F. P., FORMIGHIERI, A., SANTOS, C. A. N., FERREIRA, M. A. D., FIGUEIREDO, E. O., FERREIRA, E. J. L.
Tipo de documento: Artigo
Idioma: eng
Título da fonte: Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice)
Texto Completo: http://www.alice.cnptia.embrapa.br/alice/handle/doc/1124129
https://doi.org/10.1016/j.foreco.2020.118397
Resumo: Information regarding the spatial distribution of palm trees in tropical forests is crucial for commercial exploitation and management. However, spatially continuous knowledge of palms occurrence is scarce and difficult to obtain with conventional approaches such as field inventories. Here, we developed a new method to map Amazonian palm species at the individual tree crown (ITC) level using RGB images acquired by a low-cost unmanned aerial vehicle (UAV). Our approach is based on morphological operations performed in the score maps of palm species derived from a fully convolutional neural network model. We first constructed a labeled dataset by dividing the study area (135 ha within an old-growth Amazon forest) into 28 plots of 250 m×150 m. Then, we manually outlined all palm trees seen in RGB images with 4 cm pixels. We identified three palm species: Attalea butyracea, Euterpe precatoria and Iriartea deltoidea. We randomly selected 22 plots (80%) for training and six plots (20%) for testing. We changed the plots for training and testing to evaluate the variabilityn, in the classification accuracy and assess model generalization. Our method outperformed the average producer?s accuracy of conventional patch-wise semantic segmentation (CSS) in 4.7%. Moreover, our method correctly identified, on average, 34.7 percentage points more ITCs than CSS, which tended to merge trees that are close to each other. The producer's accuracy of A. butyracea, E. precatoria and I. deltoidea was 78.6 ± 5.5%, 8.6 ± 1.4% and 96.6 ± 3.4%, respectively. Fortunately, one of the most exploited and commercialized palm species in the Amazon (E. precatoria, a.k.a, Açaí) was mapped with the highest classification accuracy. Maps of E. precatoria derived from low-cost UAV systems can support management projects and community-based forest monitoring programs in the Amazon.
id EMBR_fd72d1cc27add3aabc9b7376fe9be4da
oai_identifier_str oai:www.alice.cnptia.embrapa.br:doc/1124129
network_acronym_str EMBR
network_name_str Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice)
repository_id_str 2154
spelling Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.PalmeiraPalm treesMapeamentoDroneAerial surveysImagem RGBDeepLabv3+Bosques lluviososMadera tropicalTeledetecciónVehículos aéreos no tripuladosFotografía aéreaEmbrapa AcreRio Branco (AC)AcreAmazônia OcidentalWestern AmazonAmazAmazonia OccidentalFloresta TropicalEspécie NativaAçaíPopulação de PlantaBiogeografiaSensoriamento RemotoAerofotogrametriaRain forestsArecaceaeEuterpe precatoriaTropical woodBiogeographyRemote sensingUnmanned aerial vehiclesAerial photographyInformation regarding the spatial distribution of palm trees in tropical forests is crucial for commercial exploitation and management. However, spatially continuous knowledge of palms occurrence is scarce and difficult to obtain with conventional approaches such as field inventories. Here, we developed a new method to map Amazonian palm species at the individual tree crown (ITC) level using RGB images acquired by a low-cost unmanned aerial vehicle (UAV). Our approach is based on morphological operations performed in the score maps of palm species derived from a fully convolutional neural network model. We first constructed a labeled dataset by dividing the study area (135 ha within an old-growth Amazon forest) into 28 plots of 250 m×150 m. Then, we manually outlined all palm trees seen in RGB images with 4 cm pixels. We identified three palm species: Attalea butyracea, Euterpe precatoria and Iriartea deltoidea. We randomly selected 22 plots (80%) for training and six plots (20%) for testing. We changed the plots for training and testing to evaluate the variabilityn, in the classification accuracy and assess model generalization. Our method outperformed the average producer?s accuracy of conventional patch-wise semantic segmentation (CSS) in 4.7%. Moreover, our method correctly identified, on average, 34.7 percentage points more ITCs than CSS, which tended to merge trees that are close to each other. The producer's accuracy of A. butyracea, E. precatoria and I. deltoidea was 78.6 ± 5.5%, 8.6 ± 1.4% and 96.6 ± 3.4%, respectively. Fortunately, one of the most exploited and commercialized palm species in the Amazon (E. precatoria, a.k.a, Açaí) was mapped with the highest classification accuracy. Maps of E. precatoria derived from low-cost UAV systems can support management projects and community-based forest monitoring programs in the Amazon.Matheus Pinheiro Ferreira, Instituto Militar de Engenharia (IME); Danilo Roberti Alves de Almeida, Universidade de São Paulo (USP); DANIEL DE ALMEIDA PAPA, CPAF-AC; Juliano Baldez Silva Minervino, Universidade Federal do Acre (Ufac); Hudson Franklin Pessoa Veras, Universidade Federal do Paraná (UFPR); Arthur Formighieri, Universidade Federal do Acre (Ufac); Caio Alexandre Nascimento Santos, Bolsista Embrapa Acre; Marcio Aurélio Dantas Ferreira, Fundação de Tecnologia do Estado do Acre (Funtac); EVANDRO ORFANO FIGUEIREDO, CPAF-AC; Evandro José Linhares Ferreira, Instituto Nacional de Pesquisas da Amazônia (Inpa).FERREIRA, M. P.ALMEIDA, D. R. A. dePAPA, D. de A.MINERVINO, J. B. S.VERAS, H. F. P.FORMIGHIERI, A.SANTOS, C. A. N.FERREIRA, M. A. D.FIGUEIREDO, E. O.FERREIRA, E. J. L.2020-08-01T11:12:33Z2020-08-01T11:12:33Z2020-07-312020info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articleForest Ecology and Management, v. 475, n. 118397, p. 1-11, 2020.0378-1127http://www.alice.cnptia.embrapa.br/alice/handle/doc/1124129https://doi.org/10.1016/j.foreco.2020.118397enginfo:eu-repo/semantics/openAccessreponame:Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice)instname:Empresa Brasileira de Pesquisa Agropecuária (Embrapa)instacron:EMBRAPA2020-08-01T11:12:41Zoai:www.alice.cnptia.embrapa.br:doc/1124129Repositório InstitucionalPUBhttps://www.alice.cnptia.embrapa.br/oai/requestopendoar:21542020-08-01T11:12:41falseRepositório InstitucionalPUBhttps://www.alice.cnptia.embrapa.br/oai/requestcg-riaa@embrapa.bropendoar:21542020-08-01T11:12:41Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice) - Empresa Brasileira de Pesquisa Agropecuária (Embrapa)false
dc.title.none.fl_str_mv Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.
title Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.
spellingShingle Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.
FERREIRA, M. P.
Palmeira
Palm trees
Mapeamento
Drone
Aerial surveys
Imagem RGB
DeepLabv3+
Bosques lluviosos
Madera tropical
Teledetección
Vehículos aéreos no tripulados
Fotografía aérea
Embrapa Acre
Rio Branco (AC)
Acre
Amazônia Ocidental
Western Amazon
Amaz
Amazonia Occidental
Floresta Tropical
Espécie Nativa
Açaí
População de Planta
Biogeografia
Sensoriamento Remoto
Aerofotogrametria
Rain forests
Arecaceae
Euterpe precatoria
Tropical wood
Biogeography
Remote sensing
Unmanned aerial vehicles
Aerial photography
title_short Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.
title_full Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.
title_fullStr Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.
title_full_unstemmed Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.
title_sort Individual tree detection and species classification of Amazonian palms using UAV images and deep learning.
author FERREIRA, M. P.
author_facet FERREIRA, M. P.
ALMEIDA, D. R. A. de
PAPA, D. de A.
MINERVINO, J. B. S.
VERAS, H. F. P.
FORMIGHIERI, A.
SANTOS, C. A. N.
FERREIRA, M. A. D.
FIGUEIREDO, E. O.
FERREIRA, E. J. L.
author_role author
author2 ALMEIDA, D. R. A. de
PAPA, D. de A.
MINERVINO, J. B. S.
VERAS, H. F. P.
FORMIGHIERI, A.
SANTOS, C. A. N.
FERREIRA, M. A. D.
FIGUEIREDO, E. O.
FERREIRA, E. J. L.
author2_role author
author
author
author
author
author
author
author
author
dc.contributor.none.fl_str_mv Matheus Pinheiro Ferreira, Instituto Militar de Engenharia (IME); Danilo Roberti Alves de Almeida, Universidade de São Paulo (USP); DANIEL DE ALMEIDA PAPA, CPAF-AC; Juliano Baldez Silva Minervino, Universidade Federal do Acre (Ufac); Hudson Franklin Pessoa Veras, Universidade Federal do Paraná (UFPR); Arthur Formighieri, Universidade Federal do Acre (Ufac); Caio Alexandre Nascimento Santos, Bolsista Embrapa Acre; Marcio Aurélio Dantas Ferreira, Fundação de Tecnologia do Estado do Acre (Funtac); EVANDRO ORFANO FIGUEIREDO, CPAF-AC; Evandro José Linhares Ferreira, Instituto Nacional de Pesquisas da Amazônia (Inpa).
dc.contributor.author.fl_str_mv FERREIRA, M. P.
ALMEIDA, D. R. A. de
PAPA, D. de A.
MINERVINO, J. B. S.
VERAS, H. F. P.
FORMIGHIERI, A.
SANTOS, C. A. N.
FERREIRA, M. A. D.
FIGUEIREDO, E. O.
FERREIRA, E. J. L.
dc.subject.por.fl_str_mv Palmeira
Palm trees
Mapeamento
Drone
Aerial surveys
Imagem RGB
DeepLabv3+
Bosques lluviosos
Madera tropical
Teledetección
Vehículos aéreos no tripulados
Fotografía aérea
Embrapa Acre
Rio Branco (AC)
Acre
Amazônia Ocidental
Western Amazon
Amaz
Amazonia Occidental
Floresta Tropical
Espécie Nativa
Açaí
População de Planta
Biogeografia
Sensoriamento Remoto
Aerofotogrametria
Rain forests
Arecaceae
Euterpe precatoria
Tropical wood
Biogeography
Remote sensing
Unmanned aerial vehicles
Aerial photography
topic Palmeira
Palm trees
Mapeamento
Drone
Aerial surveys
Imagem RGB
DeepLabv3+
Bosques lluviosos
Madera tropical
Teledetección
Vehículos aéreos no tripulados
Fotografía aérea
Embrapa Acre
Rio Branco (AC)
Acre
Amazônia Ocidental
Western Amazon
Amaz
Amazonia Occidental
Floresta Tropical
Espécie Nativa
Açaí
População de Planta
Biogeografia
Sensoriamento Remoto
Aerofotogrametria
Rain forests
Arecaceae
Euterpe precatoria
Tropical wood
Biogeography
Remote sensing
Unmanned aerial vehicles
Aerial photography
description Information regarding the spatial distribution of palm trees in tropical forests is crucial for commercial exploitation and management. However, spatially continuous knowledge of palms occurrence is scarce and difficult to obtain with conventional approaches such as field inventories. Here, we developed a new method to map Amazonian palm species at the individual tree crown (ITC) level using RGB images acquired by a low-cost unmanned aerial vehicle (UAV). Our approach is based on morphological operations performed in the score maps of palm species derived from a fully convolutional neural network model. We first constructed a labeled dataset by dividing the study area (135 ha within an old-growth Amazon forest) into 28 plots of 250 m×150 m. Then, we manually outlined all palm trees seen in RGB images with 4 cm pixels. We identified three palm species: Attalea butyracea, Euterpe precatoria and Iriartea deltoidea. We randomly selected 22 plots (80%) for training and six plots (20%) for testing. We changed the plots for training and testing to evaluate the variabilityn, in the classification accuracy and assess model generalization. Our method outperformed the average producer?s accuracy of conventional patch-wise semantic segmentation (CSS) in 4.7%. Moreover, our method correctly identified, on average, 34.7 percentage points more ITCs than CSS, which tended to merge trees that are close to each other. The producer's accuracy of A. butyracea, E. precatoria and I. deltoidea was 78.6 ± 5.5%, 8.6 ± 1.4% and 96.6 ± 3.4%, respectively. Fortunately, one of the most exploited and commercialized palm species in the Amazon (E. precatoria, a.k.a, Açaí) was mapped with the highest classification accuracy. Maps of E. precatoria derived from low-cost UAV systems can support management projects and community-based forest monitoring programs in the Amazon.
publishDate 2020
dc.date.none.fl_str_mv 2020-08-01T11:12:33Z
2020-08-01T11:12:33Z
2020-07-31
2020
dc.type.driver.fl_str_mv info:eu-repo/semantics/publishedVersion
info:eu-repo/semantics/article
format article
status_str publishedVersion
dc.identifier.uri.fl_str_mv Forest Ecology and Management, v. 475, n. 118397, p. 1-11, 2020.
0378-1127
http://www.alice.cnptia.embrapa.br/alice/handle/doc/1124129
https://doi.org/10.1016/j.foreco.2020.118397
identifier_str_mv Forest Ecology and Management, v. 475, n. 118397, p. 1-11, 2020.
0378-1127
url http://www.alice.cnptia.embrapa.br/alice/handle/doc/1124129
https://doi.org/10.1016/j.foreco.2020.118397
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.source.none.fl_str_mv reponame:Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice)
instname:Empresa Brasileira de Pesquisa Agropecuária (Embrapa)
instacron:EMBRAPA
instname_str Empresa Brasileira de Pesquisa Agropecuária (Embrapa)
instacron_str EMBRAPA
institution EMBRAPA
reponame_str Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice)
collection Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice)
repository.name.fl_str_mv Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice) - Empresa Brasileira de Pesquisa Agropecuária (Embrapa)
repository.mail.fl_str_mv cg-riaa@embrapa.br
_version_ 1794503494254723072