• Login
    View Item 
    •   DSpace Home
    • BIBLIOTECA DIGITAL: Teses & Dissertações
    • 40001016002P6 Programa de Pós-Graduação em Ciências Geodésicas
    • Dissertações
    • View Item
    •   DSpace Home
    • BIBLIOTECA DIGITAL: Teses & Dissertações
    • 40001016002P6 Programa de Pós-Graduação em Ciências Geodésicas
    • Dissertações
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Avaliação multitemporal de alteração da cobertura vegetal equatoriana por meio de imagens landsat e redes neurais artificiais

    Thumbnail
    View/Open
    R - D - LUCIA VERONICA JARAMILLO TOLEDO.pdf (11.47Mb)
    Date
    2015
    Author
    Jaramillo Toledo, Lucía Verónica
    Metadata
    Show full item record
    Subject
    Geodésia
    Redes neurais (Computação)
    Sensoriamento remoto
    Landsat (Satelites)
    Teses
    xmlui.dri2xhtml.METS-1.0.item-type
    Dissertação
    Abstract
    Resumo: A classificação e detecção de mudanças da cobertura vegetal em grandes áreas pode ser realizada mediante o uso das tecnologias associadas ao Sensoriamento Remoto. Imagens de satélite de média e alta resolução espacial e espectral contrastadas com informação coletada em campo, constituem ferramentas indispensáveis quando o objetivo de projetos e pesquisas é a discriminação entre distintos tipos de cobertura vegetal e suas variações temporais. As tarefas antes mencionadas de forma geral envolvem grandes investimentos em recursos humanos e econômicos, os resultados nem sempre são satisfatórios com a aplicação de algoritmos e técnicas orientadas à automatização do processo de classificação das distintas feições contidas na imagem, e em geral é requerida uma etapa de interpretação visual. As técnicas de Inteligência Artificial (IA), no caso específico das Redes Neurais Artificiais (RNAs) mostram ser um mecanismo adequado para realizar a classificação automática de imagens de satélite quando o objetivo é identificar mudanças na cobertura vegetal, sendo que o desempenho da técnica depende fundamentalmente do conjunto de dados de treinamento e da arquitetura da RNA. No presente trabalho, o método é testado em uma região da Amazônia equatoriana, o uso da técnica de RNAs para a detecção de alterações mediante imagens de satélite, apresentou melhores resultados do que aqueles obtidos com o método tradicional de máxima verossimilhança em termos de acurácia global e o coeficiente Kappa (superior a 0,8). Em geral, as alterações detectadas na cobertura vegetal natural estão associadas a fenómenos relacionadas com a expansão de áreas agrícolas e urbanas. Palavras-chave: Redes Neurais Artificiais, Detecção de mudanças, Classificação, Segmentação.
     
    Abstract: The classification and change detection in vegetation cover over large areas can be accomplished using technologies associated with Remote Sensing. Satellite images of middle and high spatial and spectral resolution contrasted with information collected in the field are indispensable tools when the purpose of projects and research is discrimination between different types of vegetation and their temporal variations. The tasks mentioned above generally involve large investments in human and economic resources, the results are not always satisfactory with the application of algorithms and techniques aimed at the automation of the classification process of the different features contained in the image, and is usually required one-step visual interpretation. The Artificial Intelligence (AI) techniques, in the specific case of Artificial Neural Networks (ANN) prove to be an adequate mechanism for automating the satellite image classification when the goal is to identify changes in vegetation cover, and the technical performance depends the set of RNA training data. In this paper the method is tested in a region of the Ecuadorian Amazon, getting in terms of agreement, achieve a kappa index above 0,8. Key-words: Artificial Neural Networks, Change Detection, Classification, Segmentation.
     
    URI
    http://hdl.handle.net/1884/53292
    Collections
    • Dissertações [250]

    DSpace software copyright © 2002-2016  DuraSpace
    Contact Us | Send Feedback
    Theme by 
    Atmire NV
     

     

    Browse

    All of DSpaceCommunities & CollectionsBy Issue DateAuthorsTitlesSubjectsxmlui.ArtifactBrowser.Navigation.browse_typeThis CollectionBy Issue DateAuthorsTitlesSubjectsxmlui.ArtifactBrowser.Navigation.browse_type

    My Account

    LoginRegister

    Statistics

    View Usage Statistics

    DSpace software copyright © 2002-2016  DuraSpace
    Contact Us | Send Feedback
    Theme by 
    Atmire NV