%0 Conference Proceedings
%4 dpi.inpe.br/sbsr@80/2008/11.16.03.32
%2 dpi.inpe.br/sbsr@80/2008/11.16.03.32.28
%@isbn 978-85-17-00044-7
%T Segmentação de imagens multiespectrais utilizando-se o atributo matiz
%D 2009
%A Aboud Neta, Sumaia Resegue,
%A Dutra, Luciano Vieira,
%A Prado, Bruno Rodrigues do,
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@electronicmailaddress sumaia@dpi.inpe.br
%@electronicmailaddress dutra@dpi.inpe.br
%@electronicmailaddress prado@dsr.inpe.br
%E Epiphanio, José Carlos Neves,
%E Galvão, Lênio Soares,
%B Simpósio Brasileiro de Sensoriamento Remoto, 14 (SBSR)
%C Natal
%8 25-30 abr. 2009
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%J São José dos Campos
%P 6719-6726
%S Anais
%1 Instituto Nacional de Pesquisas Espaciais (INPE)
%K Hue, segmentation, color composit, RGB, remote sensing, matiz, segmentação, composição.
%X This paper had as objective, evaluation of segmentation algorithms and classification of hue images from color composit. The advantage of working with this representation of color, is the possibility of separating the intensity of hue and saturation information. In this context, this article discusses the segmentation of multispectral images based on hue. It generated a number of results of the classification of scenes from the segmentation of the image of hue, based on a synthetic image and a high resolution image. As high-resolution image was used as an image IKONOS scene taking the city of Sao Jose dos Campos. Through the tests performed and the images presented, it was possible to prove the efficiency of the segment hue. This is evidenced by the good separability of targets with different conditions of illumination, as seen in high-resolution image. However, in some tests done where the light was too low, the value of nuance has become unstable, receiving values without any meaning. At all stages of testing, it was found that the average hue of the image was preserved, ie the segmentation of images filtered and grouped, their average values of hue were almost equal. Noting thus the integrity of the method used in the calculation of average value of hue.
%9 Processamento de Imagens
%@language pt
%3 6719-6726.pdf