%0 Conference Proceedings
%@isbn 978-85-17-00066-9 (Internet)
%@isbn 978-85-17-00065-2 (DVD)
%F 1444
%T Mineração de dados para análise da cobertura florestal
%D 2013
%A Saito, Nathália Suemi,
%A Arguello, Fernanda Viana Paiva,
%A Moreira, Maurício Alves,
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%E Epiphanio, José Carlos Neves,
%E Galvão, Lênio Soares,
%B Simpósio Brasileiro de Sensoriamento Remoto, 16 (SBSR)
%C Foz do Iguaçu
%8 13-18 abr. 2013
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%J São José dos Campos
%P 2400-2407
%S Anais
%1 Instituto Nacional de Pesquisas Espaciais (INPE)
%X The landscape ecology metrics associated with spectral and space data mining can be used to increase the potential of the analysis and applications of remote sensing data becoming an important tool for decision making. The present study sought to use data mining techniques and metrics of landscape ecology to classify and quantify the different types of vegetation present in São Luis do Paraitinga city, Sao Paulo, Brazil. The images went through object-oriented analysis through the plugin GeoDMA to obtain spectral and spatial data. This information was used to classify classes by decision trees. Eucalyptus and Forest fragment areas represented 8.6% and 36,1% of the total area, respectively. The decision tree generated by the classification algorithm was used to obtain the map of forest cover. The classification by decision tree showed kappa of 0.80, indicating little confusion. The results indicate the importance of the forest sector and contribute in studies to contain the impacts and problems caused by the expansion of eucalyptus plantations in the municipality. The generation of classifications by the method of data mining metrics associated with landscape ecology proved to be an affordable and reliable tool to extract the spatial and spectral data with remote sensing techniques. The method was efficient for the separation of forest classes and the metrics used were important to better understand the objects inserted in the landscape and the pressures they suffer.
%9 Classificação e Mineração de Dados
%@language pt
%3 p1444.pdf