• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 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

AnÃlise da dinÃmica espacial do ecossistema manguezal com abordagem metodolÃgica orientada a objeto / Analysis of the spatial dynamics of the mangrove ecosystem approach to object-oriented methodology

Paulo Roberto Lopes Thiers 27 March 2013 (has links)
Para o conhecimento da dinÃmica espacial das Ãreas de manguezais dos em dez complexos estuarinos da costa oeste cearense foram utilizadas imagens orbitais com resoluÃÃo espacial de 30m e imagens coloridas com resoluÃÃo de 2m, com aplicaÃÃo de tÃcnicas de sensoriamento remoto e tecnologia de Sistemas de InformaÃÃes GeogrÃfias (SIG). Ãs imagens do sensor TM do satÃlite LANDSAT 5, observadas por perÃodo de 23 anos, e Ãs ortofotos que recobrem a Ãrea objeto, foram aplicados procedimentos de classificaÃÃo orientada a objetos, tÃcnica que traduz o Estado da Arte para a extraÃÃo de informaÃÃes espaciais. A abordagem orientada a objeto delimita corpos singulares e homogÃneos a partir das informaÃÃes espectrais dos pixels, e de caracterÃsticas relacionadas à geometria e topologia dos objetos. Esse algoritmo foi desenvolvido para classificaÃÃo em imagens com altÃssima resoluÃÃo espacial, onde nÃo hà mistura espectral: os pixels possuem caracterÃsticas semelhantes a de seus vizinhos, enquanto que imagens de baixa e mÃdia resoluÃÃo espacial, os pixels contÃem informaÃÃes espectrais correspondentes a vÃrios objetos. Embora a base teÃrica/conceitual do mÃtodo de classificaÃÃo com orientaÃÃo a objeto priorize aplicaÃÃes sobre imagens de altÃssima resoluÃÃo espacial, a pesquisa teve por objetivo desenvolver estudos sobre imagens de mÃdia resoluÃÃo espacial para a quantificaÃÃo de Ãreas ocupadas por vegetaÃÃo de mangues dos estuÃrios da Costa Oeste cearense, entre os anos de 1985 e 2008. Com resultados da classificaÃÃo orientada a objeto, produzida em imagens de mÃdia resoluÃÃo espacial, foram realizadas anÃlises da dinÃmica espacial da Ãrea ocupada pelo sistema manguezal ao longo do recorte temporal estabelecido. A anÃlise da dinÃmica espacial entre o 1 e 2 perÃodos registrou incremento correspondente à variaÃÃo percentual de 3,43%, e para o intervalo entre o 2 e 3 perÃodos foi registrada a variaÃÃo de 6,35%. Jà a variaÃÃo observada entre os valores de Ãreas de mangues entre o 1 e 3 perÃodos, que corresponde ao recorte temporal de 1985 a 2008, corresponde a aproximadamente 10%. Em complementaÃÃo aos objetivos estabelecidos, os estudos foram concentrados na aquisiÃÃo de valores relativos à quantificaÃÃo, com acurÃcia, da vegetaÃÃo do ecossistema manguezal, mas obtidos sobre imagens ortorretificadas de alta resoluÃÃo, para a comparaÃÃo com dados publicados em trabalhos de outros autores, elaborados com tÃcnicas e metodologias distintas das empregadas na pesquisa. Para a validaÃÃo dos resultados da classificaÃÃo foram utilizadas as mesmas ortofotos, obtidas no ano de 2007, como referÃncia para a verificaÃÃo da Verdade Terrestre, com vistas à garantia da qualidade geomÃtrica e posicional das feiÃÃes de bosques de mangues identificadas. Enfatize-se que os resultados apresentaram discrepÃncias aceitÃveis para os objetivos estabelecidos. Por fim, as discrepÃncias observadas nos valores referentes aos complexos estuarinos pesquisados, comparados com outros trabalhos publicados estÃo associadas Ãs diferentes metodologias e tÃcnicas empregadas. / To the knowledge of the spatial dynamics of mangrove areas of estuarine complex on the west coast of Cearà were used satellite images with spatial resolution of 30m and images colored with a resolution of 2m, with application of remote sensing technology and geographic information systems (GIS). The sensor images of LANDSAT TM 5, observed for a period of 23 years, and orthophotos that cover the area object, classification procedures were applied object-oriented technique that reflects the state of the art for extracting spatial information. The object-oriented approach delimits natural and homogeneous bodies from the spectral information of pixels, and features related to geometry and topology of objects. This algorithm was developed for classification of images with high spatial resolution, where no spectral mixing: the pixels have similar characteristics to their neighbors, while images of low and medium spatial resolution, pixels contain spectral information corresponding to multiple objects. Although the theoretical / conceptual classification method with object orientation prioritize applications on images of high spatial resolution, the research aimed to develop studies on images of medium spatial resolution for quantification of areas occupied by vegetation mangrove estuaries Coast West Ceara, between the years 1985 and 2008. With results of object-oriented classification, produced in medium spatial resolution images were analyzed with the spatial dynamics of the area occupied by the system mangrove along the established time frame. The analysis of the spatial dynamics between the 1st and 2nd periods recorded an increase corresponding to the percentage change of 3.43%, and the interval between the 2nd and 3rd periods was recorded variation of 6.35%. However, the variation observed between the values of mangrove areas between the 1st and 3rd periods, corresponding to the time frame from 1985 to 2008, the rate recorded is around 10%. Complementing the established objectives, studies were concentrated on the acquisition of relative values to quantify with accuracy, the vegetation of the mangrove ecosystem, but obtained on high-resolution orthorectified images, for comparison with data published in works of other authors, drawn up techniques and different methodologies employed in the research. For validating the classification results were the same used orthophotos, obtained in 2007, as a reference for verifying the Truth Road, with a view to ensuring the quality of geometric and positional features of mangrove forests identified. Emphasize that the results showed discrepancies acceptable to the established objectives. Finally, the discrepancies observed in the values for the complex estuarine surveyed, compared with other published works are associated with different methodologies and techniques employed.

Page generated in 0.0516 seconds