Aprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alata
Autor(a) principal: | |
---|---|
Data de Publicação: | 2020 |
Tipo de documento: | Dissertação |
Idioma: | por |
Título da fonte: | Repositório Institucional da UFMS |
Texto Completo: | https://repositorio.ufms.br/handle/123456789/3834 |
Resumo: | This dissertation proposes the investigation of a method based on deep learning applied in the classification of Dipteryx alata, popularly known as Cumbaru, which is an arboreal species of environmental interest from Mato Grosso do Sul (MS) from RGB images collected by UAV (Unmanned Aerial Vehicles). It is organized in two chapters, the first chapter presents general considerations regarding the studied species, its socio-environmental relevance and the pertinent legislation for environmental licensing and forest inventory. It also presents a brief approach to deep learning. The second chapter aimed to evaluate the RetinaNet object detection method in identifying the species of environmental interest. The collection of images was carried out in selected places with the use of UAV and a bank of images was generated with notes of the studied species; Initial experiments were carried out within the UFMS campus (Federal University of Mato Grosso do Sul) and also in nearby areas, with a focus on mapping and monitoring Dipteryx alata. The collection period was between August 2018 and December 2019. The images were divided into training, validation and testing in the proportion of 60%, 20% and 20%. The investigated approach is based on the RetinaNet object detection method, which uses annotations with surrounding rectangles. Subsequently, a test was performed with images collected in a distant place from UFMS, aiming to evaluate the generalization capacity of the method employed. Detection accuracy was around 80% for this second test area. |
id |
UFMS_c8dbeea8edb10b34b389d86ec8aecb68 |
---|---|
oai_identifier_str |
oai:repositorio.ufms.br:123456789/3834 |
network_acronym_str |
UFMS |
network_name_str |
Repositório Institucional da UFMS |
repository_id_str |
2124 |
spelling |
2021-07-15T19:00:00Z2021-09-30T19:57:00Z2020https://repositorio.ufms.br/handle/123456789/3834This dissertation proposes the investigation of a method based on deep learning applied in the classification of Dipteryx alata, popularly known as Cumbaru, which is an arboreal species of environmental interest from Mato Grosso do Sul (MS) from RGB images collected by UAV (Unmanned Aerial Vehicles). It is organized in two chapters, the first chapter presents general considerations regarding the studied species, its socio-environmental relevance and the pertinent legislation for environmental licensing and forest inventory. It also presents a brief approach to deep learning. The second chapter aimed to evaluate the RetinaNet object detection method in identifying the species of environmental interest. The collection of images was carried out in selected places with the use of UAV and a bank of images was generated with notes of the studied species; Initial experiments were carried out within the UFMS campus (Federal University of Mato Grosso do Sul) and also in nearby areas, with a focus on mapping and monitoring Dipteryx alata. The collection period was between August 2018 and December 2019. The images were divided into training, validation and testing in the proportion of 60%, 20% and 20%. The investigated approach is based on the RetinaNet object detection method, which uses annotations with surrounding rectangles. Subsequently, a test was performed with images collected in a distant place from UFMS, aiming to evaluate the generalization capacity of the method employed. Detection accuracy was around 80% for this second test area.Esta dissertação propõe a investigação de método baseado em aprendizado profundo (deep learning) aplicado na classificação de Dipteryx alata, popularmente conhecido por Cumbaru, que é uma espécie arbórea de interesse ambiental de Mato Grosso do Sul (MS) a partir de imagens RGB coletadas por VANT (Veículos Aéreos Não-Tripulados). Está organizada em dois capítulos, o primeiro capítulo apresenta considerações gerais a respeito da espécie estudada, sua relevância socioambiental e as legislações pertinentes a licenciamento ambiental e inventário florestal. Também apresenta uma breve abordagem sobre aprendizado profundo (deep learning). O segundo capítulo teve como objetivo avaliar o método de detecção de objetos RetinaNet na identificação da espécie de interesse ambiental. Realizou-se a coleta de imagens em locais selecionados com o uso de VANT e gerou-se um banco de imagens com anotações da espécie estudada; Experimentos iniciais foram realizados dentro do campus da UFMS (Universidade Federal de Mato Grosso do Sul) e também em áreas próximas, com foco no mapeamento e monitoramento de Dipteryx alata. O período de coleta foi entre agosto de 2018 e dezembro de 2019. As imagens foram divididas em treino, validação e teste na proporção de 60%, 20% e 20%. A abordagem investigada baseia-se no método de detecção de objetos RetinaNet, que utiliza anotações com retângulos envolventes. Posteriormente, realizou-se teste com imagens coletadas em local distante da UFMS, visando avaliar a capacidade de generalização do método empregado. As precisões na detecção foram em torno de 90% para a primeira área e 80% para essa segunda área de teste.Fundação Universidade Federal de Mato Grosso do SulUFMSBrasilAprendizado profundo, VANTAprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alatainfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisJose Marcato JuniorMarcio Santos Araujoinfo:eu-repo/semantics/openAccessporreponame:Repositório Institucional da UFMSinstname:Universidade Federal de Mato Grosso do Sul (UFMS)instacron:UFMSTHUMBNAILDissertação_10_12_2020_versãocorrigida.pdf.jpgDissertação_10_12_2020_versãocorrigida.pdf.jpgGenerated Thumbnailimage/jpeg1185https://repositorio.ufms.br/bitstream/123456789/3834/3/Disserta%c3%a7%c3%a3o_10_12_2020_vers%c3%a3ocorrigida.pdf.jpg7004c81c53eb231db850ddf581344cf4MD53TEXTDissertação_10_12_2020_versãocorrigida.pdf.txtDissertação_10_12_2020_versãocorrigida.pdf.txtExtracted texttext/plain66569https://repositorio.ufms.br/bitstream/123456789/3834/2/Disserta%c3%a7%c3%a3o_10_12_2020_vers%c3%a3ocorrigida.pdf.txt5f1b4b25bf1e9c3a9729ffac5858d35fMD52ORIGINALDissertação_10_12_2020_versãocorrigida.pdfDissertação_10_12_2020_versãocorrigida.pdfapplication/pdf2688263https://repositorio.ufms.br/bitstream/123456789/3834/1/Disserta%c3%a7%c3%a3o_10_12_2020_vers%c3%a3ocorrigida.pdf46241fcfb5fe42121ef66b6bc9e96446MD51123456789/38342021-09-30 15:57:00.885oai:repositorio.ufms.br:123456789/3834Repositório InstitucionalPUBhttps://repositorio.ufms.br/oai/requestri.prograd@ufms.bropendoar:21242021-09-30T19:57Repositório Institucional da UFMS - Universidade Federal de Mato Grosso do Sul (UFMS)false |
dc.title.pt_BR.fl_str_mv |
Aprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alata |
title |
Aprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alata |
spellingShingle |
Aprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alata Marcio Santos Araujo Aprendizado profundo, VANT |
title_short |
Aprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alata |
title_full |
Aprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alata |
title_fullStr |
Aprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alata |
title_full_unstemmed |
Aprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alata |
title_sort |
Aprendizado profundo aplicado em imagens de VANT para a detecção de Dipteryx alata |
author |
Marcio Santos Araujo |
author_facet |
Marcio Santos Araujo |
author_role |
author |
dc.contributor.advisor1.fl_str_mv |
Jose Marcato Junior |
dc.contributor.author.fl_str_mv |
Marcio Santos Araujo |
contributor_str_mv |
Jose Marcato Junior |
dc.subject.por.fl_str_mv |
Aprendizado profundo, VANT |
topic |
Aprendizado profundo, VANT |
description |
This dissertation proposes the investigation of a method based on deep learning applied in the classification of Dipteryx alata, popularly known as Cumbaru, which is an arboreal species of environmental interest from Mato Grosso do Sul (MS) from RGB images collected by UAV (Unmanned Aerial Vehicles). It is organized in two chapters, the first chapter presents general considerations regarding the studied species, its socio-environmental relevance and the pertinent legislation for environmental licensing and forest inventory. It also presents a brief approach to deep learning. The second chapter aimed to evaluate the RetinaNet object detection method in identifying the species of environmental interest. The collection of images was carried out in selected places with the use of UAV and a bank of images was generated with notes of the studied species; Initial experiments were carried out within the UFMS campus (Federal University of Mato Grosso do Sul) and also in nearby areas, with a focus on mapping and monitoring Dipteryx alata. The collection period was between August 2018 and December 2019. The images were divided into training, validation and testing in the proportion of 60%, 20% and 20%. The investigated approach is based on the RetinaNet object detection method, which uses annotations with surrounding rectangles. Subsequently, a test was performed with images collected in a distant place from UFMS, aiming to evaluate the generalization capacity of the method employed. Detection accuracy was around 80% for this second test area. |
publishDate |
2020 |
dc.date.issued.fl_str_mv |
2020 |
dc.date.accessioned.fl_str_mv |
2021-07-15T19:00:00Z |
dc.date.available.fl_str_mv |
2021-09-30T19:57:00Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/masterThesis |
format |
masterThesis |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
https://repositorio.ufms.br/handle/123456789/3834 |
url |
https://repositorio.ufms.br/handle/123456789/3834 |
dc.language.iso.fl_str_mv |
por |
language |
por |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.publisher.none.fl_str_mv |
Fundação Universidade Federal de Mato Grosso do Sul |
dc.publisher.initials.fl_str_mv |
UFMS |
dc.publisher.country.fl_str_mv |
Brasil |
publisher.none.fl_str_mv |
Fundação Universidade Federal de Mato Grosso do Sul |
dc.source.none.fl_str_mv |
reponame:Repositório Institucional da UFMS instname:Universidade Federal de Mato Grosso do Sul (UFMS) instacron:UFMS |
instname_str |
Universidade Federal de Mato Grosso do Sul (UFMS) |
instacron_str |
UFMS |
institution |
UFMS |
reponame_str |
Repositório Institucional da UFMS |
collection |
Repositório Institucional da UFMS |
bitstream.url.fl_str_mv |
https://repositorio.ufms.br/bitstream/123456789/3834/3/Disserta%c3%a7%c3%a3o_10_12_2020_vers%c3%a3ocorrigida.pdf.jpg https://repositorio.ufms.br/bitstream/123456789/3834/2/Disserta%c3%a7%c3%a3o_10_12_2020_vers%c3%a3ocorrigida.pdf.txt https://repositorio.ufms.br/bitstream/123456789/3834/1/Disserta%c3%a7%c3%a3o_10_12_2020_vers%c3%a3ocorrigida.pdf |
bitstream.checksum.fl_str_mv |
7004c81c53eb231db850ddf581344cf4 5f1b4b25bf1e9c3a9729ffac5858d35f 46241fcfb5fe42121ef66b6bc9e96446 |
bitstream.checksumAlgorithm.fl_str_mv |
MD5 MD5 MD5 |
repository.name.fl_str_mv |
Repositório Institucional da UFMS - Universidade Federal de Mato Grosso do Sul (UFMS) |
repository.mail.fl_str_mv |
ri.prograd@ufms.br |
_version_ |
1807552844976357376 |