• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Classificador neural h?brido para imagens obtidas por sensoriamento remoto

Lima, Alexandre Gomes de 12 August 2011 (has links)
Made available in DSpace on 2015-03-03T15:07:33Z (GMT). No. of bitstreams: 1 AlexandreGL_DISSERT.pdf: 5013567 bytes, checksum: e16408257f23b984754d0a91e2a173b4 (MD5) Previous issue date: 2011-08-12 / Remote sensing is one technology of extreme importance, allowing capture of data from the Earth's surface that are used with various purposes, including, environmental monitoring, tracking usage of natural resources, geological prospecting and monitoring of disasters. One of the main applications of remote sensing is the generation of thematic maps and subsequent survey of areas from images generated by orbital or sub-orbital sensors. Pattern classification methods are used in the implementation of computational routines to automate this activity. Artificial neural networks present themselves as viable alternatives to traditional statistical classifiers, mainly for applications whose data show high dimensionality as those from hyperspectral sensors. This work main goal is to develop a classiffier based on neural networks radial basis function and Growing Neural Gas, which presents some advantages over using individual neural networks. The main idea is to use Growing Neural Gas's incremental characteristics to determine the radial basis function network's quantity and choice of centers in order to obtain a highly effective classiffier. To demonstrate the performance of the classiffier three studies case are presented along with the results. / O sensoriamento remoto de uma tecnologia de extrema import?ncia na atualidade, permitindo a capta??o de dados da superf?cie terrestre que s?o utilizados com diversas finalidades, entre as quais, fiscaliza??o ambiental, acompanhamento de uso dos recursos naturais, prospec??ao geol?gica e monitoramento de cat?strofes. Uma das aplica??es principais do sensoriamento remoto ? a gera??o de mapas tem?ticos e posterior levantamento de ?reas a partir de imagens geradas por sensores orbitais ou sub-orbitais. M?todos de classica??o de padr?es s?o utilizados na implementa??o de rotinas computacionais que automatizem essa atividade. As redes neurais artificiais apresentam-se como m?todos alternativos vi?veis aos classicadores estat?sticos tradicionais, principalmente em aplica??es cujos dados apresentem alta dimensionalidade como os provenientes de sensores hiperespectrais. Este trabalho tem como objetivo principal desenvolver um classicador baseado nas redes neurais de fun??o de base radial e Growing Neural Gas e que apresenta algumas vantagens em rela??o ? utiliza??o individual de redes neurais. A id?ia principal ? utilizar as caracter?sticas incrementais da rede Growing Neural Gas para determinar a quantidade e a escolha de centros da rede de fun??o de base radial com o intuito de obter um classificador altamente ecaz. Para atestar o desempenho do classicador s?o apresentados tr?s estudos de caso juntamente com os resultados obtidos

Page generated in 0.2092 seconds