Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais
Na minha lista:
| Autor principal: | |
|---|---|
| Data de Publicação: | 2024 |
| Formato: | Tese |
| Idioma: | por |
| Fonte: | Repositório Institucional da UFG |
| Texto Completo: | http://repositorio.bc.ufg.br/tede/handle/tede/13788 |
Resumo: | The classification of different crop varieties still faces significant challenges due to their similar spectral characteristics. To address this issue, the integration of remote sensing techniques with deep learning methods offers a promising solution by analyzing pixel-level data based on spectral bands, band combinations, and vegetation indices. In this study, we developed a cross-deep neural network methodology, referred to as DCN-S, with a case study focused on the classification of sugarcane varieties. The methodology was applied to remote sensing data from cultivation areas in the state of Goiás, Brazil, collected between 2019 and 2021. The DCN-S model was compared with traditional classifiers, such as k-Nearest Neighbors (kNN), Support Vector Machines (SVM), and Random Forest, as well as other neural network configurations. The results indicated that the DCN-S model achieved competitive accuracy in validation scenarios, including temporal variety considerations when compared to other studies in the literature. Moreover, the model excelled in classifying varieties without requiring the separation of developmental stages, surpassing traditional methods. Performance improvements were further observed after applying a voting process. Finally, this work’s main contributions include developing an approach for classifying agricultural varieties by combining deep learning with remote sensing data and validating this methodology in a practical scenario. The results highlight the potential of the DCN-S model to outperform traditional techniques, offering a tool for automated agricultural monitoring |
| _version_ | 1871426414659502080 |
|---|---|
| author | Kai, Priscila Marques |
| author_browse | Kai, Priscila Marques |
| author_facet | Kai, Priscila Marques |
| author_role | author |
| bitstream.checksum.fl_str_mv | 8a4605be74aa9ea9d79846c1fba20a33 4460e5956bc1d1639be9ae6146a50347 3eeffcff340ae25bdd552970769dc8f4 |
| bitstream.checksumAlgorithm.fl_str_mv | MD5 MD5 MD5 |
| bitstream.url.fl_str_mv | http://repositorio.bc.ufg.br/tede/bitstreams/0f84b620-11c4-4bdb-bc8f-b45a61bedaa2/download http://repositorio.bc.ufg.br/tede/bitstreams/f1b3cc0e-0c86-41d0-95e7-01bace32c9ab/download http://repositorio.bc.ufg.br/tede/bitstreams/c4f19941-a210-49da-8e17-da028d472ee9/download |
| collection | Repositório Institucional da UFG |
| contributor_str_mv | Costa, Ronaldo Martins da Oliveira, Bruna Mendes de Costa, Ronaldo Martins da Soares, Fabrízzio Alphonsus Alves de Melo Nunes Leitão Júnior, Plínio de Sá Arraut, Eduardo Moraes Costa, Kelton Augusto Pontara da |
| dc.contributor.advisor-co1.fl_str_mv | Oliveira, Bruna Mendes de |
| dc.contributor.advisor1.fl_str_mv | Costa, Ronaldo Martins da |
| dc.contributor.advisor1Lattes.fl_str_mv | http://lattes.cnpq.br/7080590204832262 |
| dc.contributor.author.fl_str_mv | Kai, Priscila Marques |
| dc.contributor.authorLattes.fl_str_mv | http://lattes.cnpq.br/8210180026970752 |
| dc.contributor.referee1.fl_str_mv | Costa, Ronaldo Martins da |
| dc.contributor.referee2.fl_str_mv | Soares, Fabrízzio Alphonsus Alves de Melo Nunes |
| dc.contributor.referee3.fl_str_mv | Leitão Júnior, Plínio de Sá |
| dc.contributor.referee4.fl_str_mv | Arraut, Eduardo Moraes |
| dc.contributor.referee5.fl_str_mv | Costa, Kelton Augusto Pontara da |
| dc.date.accessioned.fl_str_mv | 2025-01-21T17:18:12Z |
| dc.date.available.fl_str_mv | 2025-01-21T17:18:12Z |
| dc.date.issued.fl_str_mv | 2024-11-06 |
| dc.identifier.citation.fl_str_mv | KAI, P. M. Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais. 103 f. 2024. Tese (Doutorado em Ciência da Computação) - Instituto de Informática, Universidade Federal de Goiás, Goiânia, 2024. |
| dc.identifier.uri.fl_str_mv | http://repositorio.bc.ufg.br/tede/handle/tede/13788 |
| dc.language.iso.fl_str_mv | por |
| dc.publisher.country.fl_str_mv | Brasil |
| dc.publisher.department.fl_str_mv | Instituto de Informática - INF (RMG) |
| dc.publisher.initials.fl_str_mv | UFG |
| dc.publisher.none.fl_str_mv | Universidade Federal de Goiás |
| dc.publisher.program.fl_str_mv | Programa de Pós-graduação em Ciência da Computação (INF) |
| dc.rights.driver.fl_str_mv | Attribution-NonCommercial-NoDerivatives 4.0 International info:eu-repo/semantics/openAccess |
| dc.source.none.fl_str_mv | reponame:Repositório Institucional da UFG instname:Universidade Federal de Goiás (UFG) instacron:UFG |
| dc.subject.cnpq.fl_str_mv | CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO |
| dc.subject.eng.fl_str_mv | Sugarcane Classification Deep Learning |
| dc.subject.por.fl_str_mv | Cana-de-açúcar Classificação Aprendizado profundo |
| dc.title.alternative.eng.fl_str_mv | Deep Learning applied to pixel-level classificationof crop varieties by multispectral images |
| dc.title.none.fl_str_mv | Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais |
| dc.type.driver.fl_str_mv | info:eu-repo/semantics/doctoralThesis |
| dc.type.status.fl_str_mv | info:eu-repo/semantics/publishedVersion |
| description | The classification of different crop varieties still faces significant challenges due to their similar spectral characteristics. To address this issue, the integration of remote sensing techniques with deep learning methods offers a promising solution by analyzing pixel-level data based on spectral bands, band combinations, and vegetation indices. In this study, we developed a cross-deep neural network methodology, referred to as DCN-S, with a case study focused on the classification of sugarcane varieties. The methodology was applied to remote sensing data from cultivation areas in the state of Goiás, Brazil, collected between 2019 and 2021. The DCN-S model was compared with traditional classifiers, such as k-Nearest Neighbors (kNN), Support Vector Machines (SVM), and Random Forest, as well as other neural network configurations. The results indicated that the DCN-S model achieved competitive accuracy in validation scenarios, including temporal variety considerations when compared to other studies in the literature. Moreover, the model excelled in classifying varieties without requiring the separation of developmental stages, surpassing traditional methods. Performance improvements were further observed after applying a voting process. Finally, this work’s main contributions include developing an approach for classifying agricultural varieties by combining deep learning with remote sensing data and validating this methodology in a practical scenario. The results highlight the potential of the DCN-S model to outperform traditional techniques, offering a tool for automated agricultural monitoring |
| eu_rights_str_mv | openAccess |
| format | doctoralThesis |
| id | UFG-2_cc4f007f8fcc6a892bbd4c1906e76f50 |
| identifier_str_mv | KAI, P. M. Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais. 103 f. 2024. Tese (Doutorado em Ciência da Computação) - Instituto de Informática, Universidade Federal de Goiás, Goiânia, 2024. |
| instacron_str | UFG |
| institution | UFG |
| instname_str | Universidade Federal de Goiás (UFG) |
| language | por |
| network_acronym_str | UFG-2 |
| network_name_str | Repositório Institucional da UFG |
| oai_identifier_str | oai:repositorio.bc.ufg.br:tede/13788 |
| publishDate | 2024 |
| publishDateSort | 2024 |
| publisher.none.fl_str_mv | Universidade Federal de Goiás |
| reponame_str | Repositório Institucional da UFG |
| repository.mail.fl_str_mv | grt.bc@ufg.br |
| repository.name.fl_str_mv | Repositório Institucional da UFG - Universidade Federal de Goiás (UFG) |
| repository_id_str | oai:repositorio.bc.ufg.br:tede/1234 |
| rights_invalid_str_mv | Attribution-NonCommercial-NoDerivatives 4.0 International |
| spelling | Costa, Ronaldo Martins dahttp://lattes.cnpq.br/7080590204832262Oliveira, Bruna Mendes deCosta, Ronaldo Martins daSoares, Fabrízzio Alphonsus Alves de Melo NunesLeitão Júnior, Plínio de SáArraut, Eduardo MoraesCosta, Kelton Augusto Pontara dahttp://lattes.cnpq.br/8210180026970752Kai, Priscila Marques2025-01-21T17:18:12Z2025-01-21T17:18:12Z2024-11-06KAI, P. M. Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais. 103 f. 2024. Tese (Doutorado em Ciência da Computação) - Instituto de Informática, Universidade Federal de Goiás, Goiânia, 2024.http://repositorio.bc.ufg.br/tede/handle/tede/13788The classification of different crop varieties still faces significant challenges due to their similar spectral characteristics. To address this issue, the integration of remote sensing techniques with deep learning methods offers a promising solution by analyzing pixel-level data based on spectral bands, band combinations, and vegetation indices. In this study, we developed a cross-deep neural network methodology, referred to as DCN-S, with a case study focused on the classification of sugarcane varieties. The methodology was applied to remote sensing data from cultivation areas in the state of Goiás, Brazil, collected between 2019 and 2021. The DCN-S model was compared with traditional classifiers, such as k-Nearest Neighbors (kNN), Support Vector Machines (SVM), and Random Forest, as well as other neural network configurations. The results indicated that the DCN-S model achieved competitive accuracy in validation scenarios, including temporal variety considerations when compared to other studies in the literature. Moreover, the model excelled in classifying varieties without requiring the separation of developmental stages, surpassing traditional methods. Performance improvements were further observed after applying a voting process. Finally, this work’s main contributions include developing an approach for classifying agricultural varieties by combining deep learning with remote sensing data and validating this methodology in a practical scenario. The results highlight the potential of the DCN-S model to outperform traditional techniques, offering a tool for automated agricultural monitoringA classificação de diferentes variedades ainda enfrenta desafios significativos devido à semelhança nas características espectrais dos cultivos. Para abordar esse problema, a integração de técnicas de sensoriamento remoto com métodos de aprendizado profundo oferece uma solução promissora, analisando dados de pixeis com base em bandas espectrais, combinações de bandas e índices de vegetação. Neste trabalho, desenvolvemos uma metodologia de rede neural profunda cruzada, denominada DCN-S, com estudo de caso voltado para a classificação de variedades de cana-de-açúcar. A metodologia foi aplicada em dados de sensoriamento remoto de áreas de cultivo no estado de Goiás, coletados entre 2019 e 2021. O modelo DCN-S foi comparado com classificadores tradicionais, como kNN, SVM e Floresta Aleatória, além de outras configurações de redes neurais. Os resultados indicaram que o modelo DCN-S obteve uma acurácia competitiva em cenários de validação, incluindo a consideração de variedade temporal, quando comparado a outras investigações presentes na literatura. Além disso, o modelo se destacou na classificação de variedades sem a necessidade de separação das fases de desenvolvimento, superando os métodos tradicionais, com melhoria na performance do modelo após a aplicação de um processo de votação. Finalmente, as principais contribuições deste trabalho incluem o desenvolvimento de uma abordagem para a classificação de variedades agrícolas, combinando aprendizado profundo com dados de sensoriamento remoto, e a validação dessa metodologia em um cenário prático. Os resultados evidenciam o potencial do modelo DCN-S em superar técnicas tradicionais, oferecendo uma ferramenta para o monitoramento agrícola de forma automatizadaCoordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPESporUniversidade Federal de GoiásPrograma de Pós-graduação em Ciência da Computação (INF)UFGBrasilInstituto de Informática - INF (RMG)Attribution-NonCommercial-NoDerivatives 4.0 Internationalinfo:eu-repo/semantics/openAccessCana-de-açúcarClassificaçãoAprendizado profundoSugarcaneClassificationDeep LearningCIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAODeep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectraisDeep Learning applied to pixel-level classificationof crop varieties by multispectral imagesinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/doctoralThesisreponame:Repositório Institucional da UFGinstname:Universidade Federal de Goiás (UFG)instacron:UFGLICENSElicense.txtlicense.txttext/plain; charset=utf-81748http://repositorio.bc.ufg.br/tede/bitstreams/0f84b620-11c4-4bdb-bc8f-b45a61bedaa2/download8a4605be74aa9ea9d79846c1fba20a33MD51CC-LICENSElicense_rdflicense_rdfapplication/rdf+xml; charset=utf-8805http://repositorio.bc.ufg.br/tede/bitstreams/f1b3cc0e-0c86-41d0-95e7-01bace32c9ab/download4460e5956bc1d1639be9ae6146a50347MD52ORIGINALTese - Priscila Marques Kai - 2024.pdfTese - Priscila Marques Kai - 2024.pdfapplication/pdf10123031http://repositorio.bc.ufg.br/tede/bitstreams/c4f19941-a210-49da-8e17-da028d472ee9/download3eeffcff340ae25bdd552970769dc8f4MD53tede/137882025-01-21 14:18:12.669http://creativecommons.org/licenses/by-nc-nd/4.0/Attribution-NonCommercial-NoDerivatives 4.0 Internationalopen.accessoai:repositorio.bc.ufg.br:tede/13788http://repositorio.bc.ufg.br/tedeRepositório InstitucionalPUBhttps://repositorio.bc.ufg.br/tedeserver/oai/requestgrt.bc@ufg.bropendoar:oai:repositorio.bc.ufg.br:tede/12342025-01-21T17:18:12Repositório Institucional da UFG - Universidade Federal de Goiás (UFG)Tk9URTogUExBQ0UgWU9VUiBPV04gTElDRU5TRSBIRVJFClRoaXMgc2FtcGxlIGxpY2Vuc2UgaXMgcHJvdmlkZWQgZm9yIGluZm9ybWF0aW9uYWwgcHVycG9zZXMgb25seS4KCk5PTi1FWENMVVNJVkUgRElTVFJJQlVUSU9OIExJQ0VOU0UKCkJ5IHNpZ25pbmcgYW5kIHN1Ym1pdHRpbmcgdGhpcyBsaWNlbnNlLCB5b3UgKHRoZSBhdXRob3Iocykgb3IgY29weXJpZ2h0Cm93bmVyKSBncmFudHMgdG8gRFNwYWNlIFVuaXZlcnNpdHkgKERTVSkgdGhlIG5vbi1leGNsdXNpdmUgcmlnaHQgdG8gcmVwcm9kdWNlLAp0cmFuc2xhdGUgKGFzIGRlZmluZWQgYmVsb3cpLCBhbmQvb3IgZGlzdHJpYnV0ZSB5b3VyIHN1Ym1pc3Npb24gKGluY2x1ZGluZwp0aGUgYWJzdHJhY3QpIHdvcmxkd2lkZSBpbiBwcmludCBhbmQgZWxlY3Ryb25pYyBmb3JtYXQgYW5kIGluIGFueSBtZWRpdW0sCmluY2x1ZGluZyBidXQgbm90IGxpbWl0ZWQgdG8gYXVkaW8gb3IgdmlkZW8uCgpZb3UgYWdyZWUgdGhhdCBEU1UgbWF5LCB3aXRob3V0IGNoYW5naW5nIHRoZSBjb250ZW50LCB0cmFuc2xhdGUgdGhlCnN1Ym1pc3Npb24gdG8gYW55IG1lZGl1bSBvciBmb3JtYXQgZm9yIHRoZSBwdXJwb3NlIG9mIHByZXNlcnZhdGlvbi4KCllvdSBhbHNvIGFncmVlIHRoYXQgRFNVIG1heSBrZWVwIG1vcmUgdGhhbiBvbmUgY29weSBvZiB0aGlzIHN1Ym1pc3Npb24gZm9yCnB1cnBvc2VzIG9mIHNlY3VyaXR5LCBiYWNrLXVwIGFuZCBwcmVzZXJ2YXRpb24uCgpZb3UgcmVwcmVzZW50IHRoYXQgdGhlIHN1Ym1pc3Npb24gaXMgeW91ciBvcmlnaW5hbCB3b3JrLCBhbmQgdGhhdCB5b3UgaGF2ZQp0aGUgcmlnaHQgdG8gZ3JhbnQgdGhlIHJpZ2h0cyBjb250YWluZWQgaW4gdGhpcyBsaWNlbnNlLiBZb3UgYWxzbyByZXByZXNlbnQKdGhhdCB5b3VyIHN1Ym1pc3Npb24gZG9lcyBub3QsIHRvIHRoZSBiZXN0IG9mIHlvdXIga25vd2xlZGdlLCBpbmZyaW5nZSB1cG9uCmFueW9uZSdzIGNvcHlyaWdodC4KCklmIHRoZSBzdWJtaXNzaW9uIGNvbnRhaW5zIG1hdGVyaWFsIGZvciB3aGljaCB5b3UgZG8gbm90IGhvbGQgY29weXJpZ2h0LAp5b3UgcmVwcmVzZW50IHRoYXQgeW91IGhhdmUgb2J0YWluZWQgdGhlIHVucmVzdHJpY3RlZCBwZXJtaXNzaW9uIG9mIHRoZQpjb3B5cmlnaHQgb3duZXIgdG8gZ3JhbnQgRFNVIHRoZSByaWdodHMgcmVxdWlyZWQgYnkgdGhpcyBsaWNlbnNlLCBhbmQgdGhhdApzdWNoIHRoaXJkLXBhcnR5IG93bmVkIG1hdGVyaWFsIGlzIGNsZWFybHkgaWRlbnRpZmllZCBhbmQgYWNrbm93bGVkZ2VkCndpdGhpbiB0aGUgdGV4dCBvciBjb250ZW50IG9mIHRoZSBzdWJtaXNzaW9uLgoKSUYgVEhFIFNVQk1JU1NJT04gSVMgQkFTRUQgVVBPTiBXT1JLIFRIQVQgSEFTIEJFRU4gU1BPTlNPUkVEIE9SIFNVUFBPUlRFRApCWSBBTiBBR0VOQ1kgT1IgT1JHQU5JWkFUSU9OIE9USEVSIFRIQU4gRFNVLCBZT1UgUkVQUkVTRU5UIFRIQVQgWU9VIEhBVkUKRlVMRklMTEVEIEFOWSBSSUdIVCBPRiBSRVZJRVcgT1IgT1RIRVIgT0JMSUdBVElPTlMgUkVRVUlSRUQgQlkgU1VDSApDT05UUkFDVCBPUiBBR1JFRU1FTlQuCgpEU1Ugd2lsbCBjbGVhcmx5IGlkZW50aWZ5IHlvdXIgbmFtZShzKSBhcyB0aGUgYXV0aG9yKHMpIG9yIG93bmVyKHMpIG9mIHRoZQpzdWJtaXNzaW9uLCBhbmQgd2lsbCBub3QgbWFrZSBhbnkgYWx0ZXJhdGlvbiwgb3RoZXIgdGhhbiBhcyBhbGxvd2VkIGJ5IHRoaXMKbGljZW5zZSwgdG8geW91ciBzdWJtaXNzaW9uLgo= |
| spellingShingle | Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais Kai, Priscila Marques Cana-de-açúcar Classificação Aprendizado profundo Sugarcane Classification Deep Learning CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO |
| status_str | publishedVersion |
| title | Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais |
| title_full | Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais |
| title_fullStr | Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais |
| title_full_unstemmed | Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais |
| title_short | Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais |
| title_sort | Deep Learning aplicado à classificação em nível de pixel de variedades de culturas por imagens multiespectrais |
| topic | Cana-de-açúcar Classificação Aprendizado profundo Sugarcane Classification Deep Learning CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO |
| url | http://repositorio.bc.ufg.br/tede/handle/tede/13788 |