Uso de técnicas de sensoriamento remoto e SIG para mapeamento dos aluviões em zonas semi-áridas através de métodos de classificação automática

• Ano de defesa: 2006
• Autor(a) principal: Oliveira, Francisco de Assis Jorge
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Recursos hídricos; Sensoriamento remoto; Processamento de imagens
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/16838

Remote sensing has being widely applied in various fields of knowledge in order to produce high quality results with lower costs, computationally and financially speaking. The main goal of this research is the use remote sensing techniques combined with field surveys for the semi-automatic identification of alluvium bodies. The proposed methodology was based on adaptations of current available methodologies employed on other imagery products. This methodology is very promising since allows the interpretations of large areas with precision and very fast. The images were classified with their values in Digital Number (DN) based in samples collected from these images. In other stage of classification, the images were submitted to a pre-processing in order to correct possible errors or interferences that usually are present in these images. Among these corrections, the most important is the elimination or reduction of atmospheric interference. For the classification with revised images were used spectral signatures from collected targets in the images and also the spectral signatures from alluvium targets obtained in the field by using a spectroradiometer (equipment that registers the spectral signature the targets in wavelengths between 350 nm and 1050 nm ). These signatures are curves of continuum wavelength between 350 nm and 1050 nm, which corresponds to the spectral answer from a target. These curves can be used in the classification of land use, or in our case, in the alluvium identification. The case of study is the ITAPIUNA SUDENE-MAP and the watershed of Forquilha River located in Quixeramobim. After many classifications, evaluation of the results was carried out in order to determine the best classification for each image and then produce the alluvium map for the area of study. The results were obtained using mathematical models and visual interpretation combined.