Detecção da doença mela-das-sementes da braquiária em gramíneas forrageiras através de técnicas de sensoriamento remoto
Carregando...
Arquivos
Data
2006-02-23
Autores
Rosatti, José Carlos [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Imagens CCD/CBERS 2 nas bandas B2, B3 e B4, dos anos de 2004 e 2005, de Mirante do Paranapanema, SP (Brasil), foram transformadas em reflectância de superfície usando o modelo 5S de correção atmosférica e normalizadas radiometricamente tendo como base o trabalho desenvolvido por Hall et al. (1991). O objetivo foi caracterizar espectralmente áreas de pastagens de Brachiaria brizantha (Hoschst ex A. Rich) Stapt, em fase de florescimento, isentas e infectadas com a doença mela-das-sementes da braquiária, no sentido de possibilitar a sua detecção através da comparação entre os valores de FRBS - Fator de Reflectância Bidirecional de Superfície. Ao mesmo tempo, objetivou-se também, avaliar a eficácia das imagens CCD/CBERS-2 para a obtenção de respostas espectrais de pastagens. Os valores de reflectância de superfície foram comparados e analisados através da variação espectral existente entre as áreas sadias e doentes e os resultados indicaram que as principais diferenças foram a diminuição da reflectância na banda B3 e o aumento da reflectância na banda B4 nas áreas doentes. Diferenças significativas também foram constatadas pela comparação entre dados obtidos de imagens índices de vegetação NDVI e RVI. A metodologia empregada com o uso de imagens CCD/CBERS-2 mostrou-se eficaz para descriminar áreas infectadas com a doença.
CCD/CBERS 2 images in the spectral bands B2, B3 and B4 of the years 2004 and 2005, from Mirante do Paranapanema, SP (Brazil), were transformed in surface reflectance images using the model 5S of atmospheric correction and radiometrically normalized tends as base the work developed by Hall et al. (1991). The objective was characterize spectrally areas of pastures of Brachiaria brizantha (Hoschst ex A. Rich) Stapt, in your flowering phase, exempt and infected with the disease mela-das-sementes da braquiária, in the sense of making possible your detection through the comparison among the FRBS - Surface of Bidirectional Reflectance Factor values. At the same time, it was aimed at to evaluate the effectiveness of the CCD/CBERS-2 images for the obtaining of spectral response of pastures. The surface of reflectance values were compared and analyzed through the existent spectral variation between the healthy and diseased areas and the results they indicated that the main differences were the decrease of the reflectance in the band B3 and the increase of the reflectance in the band B4 in the diseased areas. Significant differences were also verified by the comparison among obtained of vegetation indexes NDVI and RVI data. The methodology used with the use of CCD/CBERS-2 images it was shown effective for discrimination the areas infected with the disease.
CCD/CBERS 2 images in the spectral bands B2, B3 and B4 of the years 2004 and 2005, from Mirante do Paranapanema, SP (Brazil), were transformed in surface reflectance images using the model 5S of atmospheric correction and radiometrically normalized tends as base the work developed by Hall et al. (1991). The objective was characterize spectrally areas of pastures of Brachiaria brizantha (Hoschst ex A. Rich) Stapt, in your flowering phase, exempt and infected with the disease mela-das-sementes da braquiária, in the sense of making possible your detection through the comparison among the FRBS - Surface of Bidirectional Reflectance Factor values. At the same time, it was aimed at to evaluate the effectiveness of the CCD/CBERS-2 images for the obtaining of spectral response of pastures. The surface of reflectance values were compared and analyzed through the existent spectral variation between the healthy and diseased areas and the results they indicated that the main differences were the decrease of the reflectance in the band B3 and the increase of the reflectance in the band B4 in the diseased areas. Significant differences were also verified by the comparison among obtained of vegetation indexes NDVI and RVI data. The methodology used with the use of CCD/CBERS-2 images it was shown effective for discrimination the areas infected with the disease.
Descrição
Palavras-chave
Cartografia, Pastagens, Caracterização espectral, Imagens CCD, Spectral characterization, Pastures, CDD images
Como citar
ROSATTI, José Carlos. Detecção da doença mela-das-sementes da braquiária em gramíneas forrageiras através de técnicas de sensoriamento remoto. 2006. 163 f. Dissertação (Mestrado) - Universidade Estadual Paulista, Faculdade de Ciências e Tecnologia, 2006.