%0 Conference Proceedings
%4 sid.inpe.br/marte2/2017/10.27.16.33
%2 sid.inpe.br/marte2/2017/10.27.16.33.01
%@isbn 978-85-17-00088-1
%F 59997
%T Classificação da Cobertura do Solo por meio do Uso de Imagem Worldview-2 e o Sistema InterIMAGE
%D 2017
%A Ganem, Khalil Ali,
%A Baptista, Gustavo Macedo de Mello,
%@electronicmailaddress khalilganem@hotmail.com
%E Gherardi, Douglas Francisco Marcolino,
%E Aragão, Luiz Eduardo Oliveira e Cruz de,
%B Simpósio Brasileiro de Sensoriamento Remoto, 18 (SBSR)
%C Santos
%8 28-31 maio 2017
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%J São José dos Campos
%P 7907-7914
%S Anais
%1 Instituto Nacional de Pesquisas Espaciais (INPE)
%X The object-based image analysis (OBIA) systems have been widely used in the classification of use and coverage of urban areas, where it is possible to classify features of the terrestrial surface with a higher level of precision. The advance of remote sensing technologies has opened space for the emergence of multispectral sensor systems with higher resolutions than those already available in the market. However, these improvements apply to spatial, temporal and radiometric resolutions, not covering the spectral resolution, which makes it difficult to discriminate the various targets in the urban scenario. In this context, this study uses the InterIMAGE system alongside WorldView-2 satellite imagery and data mining techniques to classify the land cover in an urban portion of Brasilia-Brazil considering a total of 10 classes, such as pavements, different types of construction roofs, vegetation cover, water and bare soil. Despite the spectral mixing among some classes such as asphalt and concrete roof, and tiled roof and bare soil, the final result demonstrated a very satisfactory and effective classification, achieving approximately an overall accuracy of 89% and a Kappa Accuracy Index of 85%. Besides showing the efficiency of the automatic mapping of land cover, this study also explored the main functions of InterIMAGE.
%9 Classificação e mineração de dados
%@language pt
%3 59997.pdf