%0 Conference Proceedings
%4 dpi.inpe.br/marte/2011/07.05.14.06
%2 dpi.inpe.br/marte/2011/07.05.14.06.28
%@isbn 978-85-17-00056-0 (Internet)
%@isbn 978-85-17-00057-7 (DVD)
%T Classificação de imagens TM5-Landsat com análise orientada a objeto para a verificação da expansão urbana de Bueno Brandão, MG
%D 2011
%A Boulomytis, Vassiliki Terezinha Galvão,
%A Alves, Cláudia Durand,
%@affiliation Instituto Nacional de Pesquisas Espaciais - INPE
%@affiliation Instituto Nacional de Pesquisas Espaciais - INPE
%@electronicmailaddress likitgb@dsr.inpe.br
%@electronicmailaddress durand@dsr.inpe.br
%E Epiphanio, José Carlos Neves,
%E Galvão, Lênio Soares,
%B Simpósio Brasileiro de Sensoriamento Remoto, 15 (SBSR).
%C Curitiba
%8 30 abr. - 5 maio 2011
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%J São José dos Campos
%P 1036-1042
%S Anais
%1 Instituto Nacional de Pesquisas Espaciais (INPE)
%K Medium spatial resolution images, detection of urban areas, object-based approach, imagens de média resolução espacial, detecção de áreas urbanas, abordagem orientada a objeto.
%X In this study orbital remote sensing data were processed in a GIS for the verification of urban expansion between 2000 and 2007 in the county of Bueno Brandão, state of Minas Gerais. The reason for choosing this period was that in the last decade, there was a growth of rural evasion in the city, due to the available governmental habitation programs. Although there was a significant and visible expansion, there was no available data of the current area covering the urban settlements. The classification of TM5-Landsat images was done by the use of an object-based approach, applied in the the Definiens 7 software. Spectral, relational and textural attributes were used for the definition of classes, including the Normalized Difference Vegetation Index (NDVI), which allows the separability of the vegetation class, and also, the textural homogeneity, which was used through specific fuzzy functions for the urban and bare soil classes, making it possible to reduce the confusion in the process of classifying medium spatial resolution images. The data integration was done by the use of SPRING software, version 5.1.5.The validation of the urban area classifications was done by the use of Kappa statistics, obtaining the indexes of 0.72 and 0.86 for the 2000 and 2007 maps. The methodology approach of this study was able to show the relevance in using this kind of classification technique as it may be systematized with low cost in short-time periods, providing good quality results, in order to guide the necessary procedures toward urban planning.
%9 Análise Urbana
%@language pt
%3 p0586.pdf