%0 Conference Proceedings
%4 sid.inpe.br/marte2/2015/06.15.15.04.06
%2 sid.inpe.br/marte2/2015/06.15.15.04.07
%@isbn 978-85-17-0076-8
%F 459
%T Caracterização espectral da floresta e áreas alagadas em imagens Worldview-2
%D 2015
%A Abreu, Marcelo Bueno de,
%A Amaral, Felipe Gonçalves,
%A Cruz, Carla Bernadete Madureira,
%@electronicmailaddress buenodeabreu@gmail.com
%E Gherardi, Douglas Francisco Marcolino,
%E Aragão, Luiz Eduardo Oliveira e Cruz de,
%B Simpósio Brasileiro de Sensoriamento Remoto, 17 (SBSR)
%C João Pessoa
%8 25-29 abr. 2015
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%J São José dos Campos
%P 2238-2245
%S Anais
%1 Instituto Nacional de Pesquisas Espaciais (INPE)
%X High resolution images have been evolving a lot in the past few years, with a steady increase in the available image collection. This has been enabling dynamics studies in larger scales, which requires a set of tools to standardize images from different dates and extract the maximum amount of information available. Atmospheric correction and image fusion are common tools used for these purposes. This work''s goal is to make a spectral characterization of forest and wetland areas on Worldview-2 images from the Silva Jardim County in Rio de Janeiro State. Reflection values from images before and after atmospheric correction and images fused with the panchromatic band were compared against each other. Two other Vegetation Indexes were created and, together with NDVI, used for the spectral characterization. The results showed that atmospheric correction met the expectation related to the vegetation target, adjusting the reflectance values of blue band relative to the green band. For the classes differentiation both infrared bands and red edge bands showed good results. The fused images indicated subtle variations to the values. When comparing fused and non fused images, the forest areas values were slightly bigger for the fused ones, whereas on wetlands the values for fused images were slightly smaller. The NDVI showed the best results among the indexes to distinguish the forest class, with values between 0,75 and 0,8.
%9 Processamento de imagens
%@language pt
%3 p0459.pdf