Close
Metadata

1. Identity statement
Reference TypeThesis or Dissertation (Thesis)
Sitemtc-m16c.sid.inpe.br
Holder Codeisadg {BR SPINPE} ibi 8JMKD3MGPCW/3DT298S
Identifier6qtX3pFwXQZ3P8SECKy/zcpUE
Repositorysid.inpe.br/jeferson/2003/08.19.09.05
Last Update2023:01.03.18.41.36 (UTC) simone
Metadata Repositorysid.inpe.br/jeferson/2003/08.19.09.05.02
Metadata Last Update2023:01.04.04.24.50 (UTC) administrator
Secondary KeyINPE--TDI/
Citation KeyCastro:2003:DeBoNa
TitleDetecção de bordas e navegação autônoma utilizando redes neurais artificiais
Alternate TitleEdge detection and autonomous navigation using artificial neural network
CourseCAP-SPG-INPE-MCT-BR
Year2003
Secondary Date20030617
Date2003-06-02
Access Date2024, May 04
Thesis TypeDissertação (Mestrado em Computação Aplicada)
Secondary TypeTDI
Number of Pages154
Number of Files1
Size24538 KiB
2. Context
AuthorCastro, Ana Paula Abrantes
GroupCAP-SPG-INPE-MCT-BR
CommitteeFonseca, Leila Maria Garcia (presidente)
Silva, José Demisio da (orientador)
Guimarães, Lamartine Nogueira Frutuoso
Silva, Ivan Nunes da
e-Mail Addressabrantesapc@gmail.com
UniversityInstituto Nacional de Pesquisas Espaciais (INPE)
CitySão José dos Campos
History (UTC)2006-09-27 21:14:33 :: administrator -> jefferson ::
2008-05-02 16:46:58 :: jefferson -> administrator ::
2008-08-21 21:18:10 :: administrator -> jefferson ::
2009-04-28 19:26:50 :: jefferson -> carolina@sid.inpe.br ::
2009-04-30 19:56:36 :: carolina@sid.inpe.br -> marciana ::
2009-06-15 17:20:33 :: marciana -> carolina@sid.inpe.br ::
2009-06-30 14:42:47 :: carolina@sid.inpe.br -> administrator ::
2009-07-07 16:12:06 :: administrator -> carolina@sid.inpe.br ::
2009-07-09 15:49:55 :: carolina@sid.inpe.br -> administrator ::
2018-07-18 16:43:43 :: administrator -> sergio :: 2003
2019-02-18 15:48:45 :: sergio :: 2003 ->
2019-02-18 15:53:46 :: sergio -> administrator ::
2020-07-07 11:05:03 :: administrator -> sergio ::
2020-07-08 13:22:44 :: sergio -> administrator ::
2022-03-15 18:56:39 :: administrator -> simone ::
2023-01-03 18:41:37 :: simone :: -> 2023
2023-01-03 18:53:21 :: simone :: 2023 -> 2003
2023-01-03 18:54:50 :: simone -> administrator :: 2003
2023-01-04 04:24:50 :: administrator -> :: 2003
3. Content and structure
Is the master or a copy?is the master
Content Stagecompleted
Transferable1
Keywordsdetecção de bordas
navegação autônoma
redes neurais
AbstractEsta dissertação estuda o uso de Redes Neurais Artificiais na detecção de bordas em imagens e no controle de um veículo em navegação autônoma. São apresentados vários modelos de redes neurais que são estudados e testados, com o objetivo de buscar o modelo mais adequado para a tarefa de detecção de bordas, segundo critérios de desempenho que comparam as redes neurais artificiais com algoritmos tradicionais na área de visão computacional, como por exemplo, o operador de Canny. Para o controle da navegação autônoma utilizam-se modelos de redes neurais com aprendizagem supervisionada, treinadas para simular os processos envolvidos na navegação realizada por um humano. O desempenho das redes neurais na navegação é comparado com um sistema baseado em lógica nebulosa usado como base para o treinamento. São usadas diferentes imagens teste na experimentação das redes neurais no processo de detecção de bordas. No trabalho é proposta uma metodologia para medir a qualidade das imagens de borda geradas pelos operadores de redes neurais. Os resultados encontrados mostram-se promissores, com as redes neurais apresentando desempenho similar ao método de Canny. ABSTRACT: This work is about the study of Artificial Neural Networks (ANN) Systems for edge detection and robot autonomous navigation. Different ANNs are studied and tested in a search for the most adequate model for edge detection according to a performance criterion that compares the ANN based detectors to the standard algorithms available in the literature, such as Canny operator. Supervised neural network models are used to simulate the human navigation control processes in autonomous navigation. Their performances are compared to a fuzzy logic control system developed earlier, whose parameters are used in the neural network training processes. Different images are used to test the neural network edge detectors. A methodology for measuring the quality of the edge images produced by the neural network operator is proposed. The results show that the neural network operators have a performance comparable to existing Canny standard operator, and are thus promising operators for edge detection.
AreaCOMP
Arrangementurlib.net > BDMCI > Fonds > Produção pgr ATUAIS > CAP > Detecção de bordas...
doc Directory Contentaccess
source Directory Contentthere are no files
agreement Directory Content
TERMO DE DEPOSITO ASSINADO ANA PAULA ABRANTES CASTRO DOUTORADO.pdf 07/07/2020 08:46 101.5 KiB 
4. Conditions of access and use
data URLhttp://urlib.net/ibi/6qtX3pFwXQZ3P8SECKy/zcpUE
zipped data URLhttp://urlib.net/zip/6qtX3pFwXQZ3P8SECKy/zcpUE
Languagept
Target FileAna Paula Abrantes de Castro.pdf
User Groupadministrator
carolina@sid.inpe.br
simone
Visibilityshown
Copyright Licenseurlib.net/www/2012/11.12.20.35
Rightsholderoriginalauthor yes
Read Permissionallow from all
Update Permissionnot transferred
5. Allied materials
Next Higher Units8JMKD3MGPCW/3F2PHGS
Host Collectionsid.inpe.br/mtc-m18@80/2008/03.17.15.17
6. Notes
Empty Fieldsacademicdepartment affiliation archivingpolicy archivist callnumber contenttype copyholder creatorhistory descriptionlevel dissemination doi electronicmailaddress format isbn issn label lineage mark mirrorrepository nextedition notes number orcid parameterlist parentrepositories previousedition previouslowerunit progress readergroup resumeid schedulinginformation secondarymark session shorttitle sponsor subject tertiarymark tertiarytype url versiontype


Close