|
|
| Acesso ao texto completo restrito à biblioteca da Embrapa Gado de Leite. Para informações adicionais entre em contato com cnpgl.biblioteca@embrapa.br. |
Registro Completo |
Biblioteca(s): |
Embrapa Gado de Leite. |
Data corrente: |
20/05/2022 |
Data da última atualização: |
15/07/2023 |
Tipo da produção científica: |
Artigo em Periódico Indexado |
Autoria: |
FRANCO, V. R.; HOTT, M. C.; ANDRADE, R. G.; GOLIATT, L. |
Afiliação: |
VICTOR REZENDE FRANCO, Universidade Federal de Juiz de Fora; MARCOS CICARINI HOTT, CNPGL; RICARDO GUIMARAES ANDRADE, CNPGL; LEONARDO GOLIATT, Universidade Federal de Juiz de Fora. |
Título: |
Hybrid machine learning methods combined with computer vision approaches to estimate biophysical parameters of pastures. |
Ano de publicação: |
2023 |
Fonte/Imprenta: |
Evolutionary Intelligence, v. 16, p. 1271-1284, 2023. |
DOI: |
https://doi.org/10.1007/s12065-022-00736-9 |
Idioma: |
Inglês |
Conteúdo: |
With population growth, the search for technologies that enable improvements in production respecting the environment and people?s health has become an essential point for society. In this context, this paper presents a study based on computer vision techniques and Machine Learning (ML) to extract information from pastures Panicum maximum cv. BRS Zuri to assist in the management and research on pasture conditions, possibilitando a obtenção de informações da. Computer vision aproaches are used to extract biophysical parameters from images acquired orthogonally from the canopy of vegetation. The extracted information serves as input for Machine Learning (ML) methods to predict pasture height and biomass. The contribution of this paper is developing a possible new solution compared to traditional methods in the large-scale study of plant biophysical parameters, which can be laborious and costly and sometimes depend on destructive harvesting. For this, three techniques were used: Support Vector Regression, Multi-Layer Perceptron (MLP), and Least Absolute Shrinkage and Selection. In addition, the Diferential Evolution technique was used to select the best model. Thirty independent runs of the Diferential Evolution technique were performed to assess the approach?s performance. The cross-validation method results show the MLP obtained the best results reaching an average of Coefcient of Determination (R2) equal 0.496 to estimate biomass and 0.656 to estimate the pasture height. |
Palavras-Chave: |
Diferential evolution; Evolução diferencial; Evolutionary model selection; Machine learning; Modelo evolutivo; Parâmetro biofísico; Rede neural; Visão computacional. |
Thesagro: |
Pastagem. |
Thesaurus Nal: |
Computer vision; Neural networks; Pastures. |
Categoria do assunto: |
X Pesquisa, Tecnologia e Engenharia |
Marc: |
LEADER 02475naa a2200313 a 4500 001 2143284 005 2023-07-15 008 2023 bl uuuu u00u1 u #d 024 7 $ahttps://doi.org/10.1007/s12065-022-00736-9$2DOI 100 1 $aFRANCO, V. R. 245 $aHybrid machine learning methods combined with computer vision approaches to estimate biophysical parameters of pastures.$h[electronic resource] 260 $c2023 520 $aWith population growth, the search for technologies that enable improvements in production respecting the environment and people?s health has become an essential point for society. In this context, this paper presents a study based on computer vision techniques and Machine Learning (ML) to extract information from pastures Panicum maximum cv. BRS Zuri to assist in the management and research on pasture conditions, possibilitando a obtenção de informações da. Computer vision aproaches are used to extract biophysical parameters from images acquired orthogonally from the canopy of vegetation. The extracted information serves as input for Machine Learning (ML) methods to predict pasture height and biomass. The contribution of this paper is developing a possible new solution compared to traditional methods in the large-scale study of plant biophysical parameters, which can be laborious and costly and sometimes depend on destructive harvesting. For this, three techniques were used: Support Vector Regression, Multi-Layer Perceptron (MLP), and Least Absolute Shrinkage and Selection. In addition, the Diferential Evolution technique was used to select the best model. Thirty independent runs of the Diferential Evolution technique were performed to assess the approach?s performance. The cross-validation method results show the MLP obtained the best results reaching an average of Coefcient of Determination (R2) equal 0.496 to estimate biomass and 0.656 to estimate the pasture height. 650 $aComputer vision 650 $aNeural networks 650 $aPastures 650 $aPastagem 653 $aDiferential evolution 653 $aEvolução diferencial 653 $aEvolutionary model selection 653 $aMachine learning 653 $aModelo evolutivo 653 $aParâmetro biofísico 653 $aRede neural 653 $aVisão computacional 700 1 $aHOTT, M. C. 700 1 $aANDRADE, R. G. 700 1 $aGOLIATT, L. 773 $tEvolutionary Intelligence$gv. 16, p. 1271-1284, 2023.
Download
Esconder MarcMostrar Marc Completo |
Registro original: |
Embrapa Gado de Leite (CNPGL) |
|
Biblioteca |
ID |
Origem |
Tipo/Formato |
Classificação |
Cutter |
Registro |
Volume |
Status |
URL |
Voltar
|
|
Registros recuperados : 323 | |
Registros recuperados : 323 | |
|
Nenhum registro encontrado para a expressão de busca informada. |
|
|