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

Classificação fuzzy de vertentes por krigagem e TPS com agregação de regiões via diagrama de Voronoi /

Berveglieri, Adilson. January 2011 (has links)
Orientador: Messias Meneguette Júnior / Banca: João Fernando Custodio da Silva / Banca: Ricardo Luís Barbosa / Resumo: As vertentes, como superf cies inclinadas, consistem em express~oes da Geomorfologia moldadas por fatores naturais (end ogenos e ex ogenos) e pelo pr oprio homem. Suas formas determinam o uxo ou o ac umulo de agua e representam caracter sticas fundamentais para a preven c~ao e resolu c~ao de problemas associados ao relevo, tais como utiliza c~ao do solo, constru c~ao civil entre outros. A classi ca c~ao da vertente em c^oncava, convexa ou retil - nea permite a identi ca c~ao de areas conforme sua declividade. Assim, por meio de uma grade retangular regular, base do modelo digital de terreno, gera-se uma malha interpolada por fun c~oes estimadoras: thin-plate spline, que possui caracter sticas de suaviza c~ao e krigagem, que al em da suavidade tamb em considera a depend^encia espacial. Logo ap os, a classi ca c~ao e feita, obedecendo a infer^encia fuzzy baseada em fun c~oes de pertin^encia que de nem classes a partir do c alculo da inclina c~ao e da concavidade ou convexidade do terreno. Entretanto, o resultado dessa classi ca c~ao est a atrelado a resolu c~ao da malha, n~ao permitindo fazer qualquer corre c~ao pontual. Pois, pequenas areas de pouca signi c^ancia podem ser formadas, necessitando elimin a-las. Nesse sentido, para que o resultado seja ajustado, aplica-se o diagrama de Voronoi, caracterizado por sua rela c~ao de abrang^encia e proximidade, como ferramenta para agregar regi~oes anteriormente classi cadas de modo a permitir um ajuste local e tornar o resultado mais condizente com a area em estudo, quando comparada a mapas geomorfol ogicos correspondentes / Abstract: Slopes, such as inclined surfaces, consist in geomorphological expressions shaped by natural factors (endogenous and exogenous) and also by man himself. Their shapes determine the ow or accumulation of water and represent fundamental characteristics for the prevention and resolution of problems associated with relief, as land use, buildings, and others. Classi- cating slopes in concave, convex or straight allows to identi cate areas based on declivity. Thus, by regular rectangular grid which represents a digital terrain model, it generates a interpolated mesh by estimator functions: thin-plate spline, which has characteristics of smoothing, and kriging, which besides smoothing also considers spatial dependence. After that, the classi cation is realized according to fuzzy inference based on membership functions that de ne classes from the calculation of the slope and concavity or convexity of the ground. However, the classi cation depends on mesh resolution and it not allows any point correction. Once small areas with little importance can be formed requiring eliminate them. In order to adjust the result, it applies the Voronoi diagram, characterized by its comprisement and close relationship and scope, as a tool to aggregate regions previously classi ed and allow a local adjustment, that can provides a consistent result in study areas, if it was compared to the corresponding geomorphological maps / Mestre

Page generated in 0.0863 seconds