Aplicabilidade da Lógica Fuzzy para classificação do uso da terra na Bacia do Rio Tenente Amaral em Jaciara/MT

• Ano de defesa: 2011
• Autor(a) principal: Barros, Adriana Oliveira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Mato Grosso; Brasil; Instituto de Ciências Humanas e Sociais (ICHS); UFMT CUC - Cuiabá; Programa de Pós-Graduação em Geografia
• Palavras-chave em Português: Sensoriamento remoto; Classificação de uso da terra; Lógica fuzzy; CNPQ::CIENCIAS HUMANAS::GEOGRAFIA; Remote sensing; Land usage classification; Fuzzy logic
• Link de acesso: http://ri.ufmt.br/handle/1/1296

Traditional techniques of classification of multispectral images have been the customary tool for thematic mapping of land cover. Such techniques were designed to deal with phenomena that show well-defined limits and can be easily discretized. There are, however, regions of uncertainty and inaccuracies to be mapped, which require alternative techniques, such as classifications based on fuzzy approach. This work, therefore, aimed to classify the use of land in the drainage basin of the Tenente Amaral river trough a fuzzy logic algorithm, supervised classificatory technique, nonparametric, utilizing a Landsat-TM image and pre-existent thematic layers. For this, it was required the generation of a database trough data survey and generation of thematic maps with the remote sensing techniques and spatial analysis, utilizing the following software: ArcGis 9.3, Spring 5.1.7 and IDRISI Andes 15.0. The spectral signatures of agricultural crops were analyzed from field samples along with satellite image, and histograms of the thematic classes of pedologic maps and steepness of the terrain were extracted, which were used as samples in the development of the fuzzy classification. The validation of the mapping was done with the kappa index, whose value was (0.8432),according to said result, the classification can be considered good to excellent. However, the work demonstrates the need to make new tests with new data in order to have better results in the classification of land usage by fuzzy approaches.