UM ESTUDO SOBRE O USO DO MODELO DE COR RGB NO PROCESSO DE CORRESPONDÊNCIA POR CORRELAÇÃO

• Autor(a) principal: Silva, Paulo Henrique Amorim da
• Data de Publicação: 2007
• Outros Autores: Tommaselli, Antonio Maria Garcia, Galo, Mauricio, Costa, Elaine
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Boletim de Ciências Geodésicas
• Texto Completo: https://revistas.ufpr.br/bcg/article/view/8247
• topic: Image Matching, Digital Photogrammetry, Computer Vision; Correlação de Imagens, Fotogrametria Digital, Visão Computacional

Image matching is one of the most important processes in Digital Photogrammetry,since it allows the automation of several stages of the photogrammetric pipeline. Inmost of the commercial software nowadays available, the algorithms of imagecorrelation use the intensity information (gray levels), despising color information,that could be useful, if used in a suitable way, increasing the robustness of thecurrent correspondence techniques in Digital Photogrammetry. The aim of thiswork is to present a technique that uses the RGB color model in the correlationprocess, in which a correlation matrix is generated for each color channel. The traceof the covariance matrix related to the translations of the reference window is usedto predict which channel can better contribute to the result of the correlation and,with this, to properly weight the correlation coefficients. The weights to be appliedto each one of the correlations matrixes are computed adaptively, considering thecharacteristics of each image. In order to assess this methodology, experiments withreal color aerial images were accomplished and correct correlations were achievedwith the proposed technique but failed with the current techniques using only greylevel images. The results are presented and discussed, showing that the use of colorinformation increases the robustness of the correlation process.