Digital Image Rescaling Using Kriging
Image was always an important knowledge source in the majority of the domains. Today, with high digital technologies, digital image processing is an essential task for almost all the phenomena including the geophysical and environmental field.The manipulation of digital image requires very often pixel manipulation. When someone needs to study image pixels created, for example, by rescaling operation it cant be acceptable resolution reduction as a method driven by an accurate measurement goal but, alternatively, it must find the best algorithms for pixels interpolation.This paper compares a rescaled 2D digital image using different non geostatistical interpolators, such as, nearest neighbour, bilinear and bicubic with kriging estimation pretending to demonstrate the ability of kriging with an automatically adjusted variogram as a good alternative to the non geostatistical traditional algorithms for pixel interpolations.Software tools for Microsoft Windows(R) were building for process images in bmp format and rescale colour and grayscale images obtained by satellite and digital cameras. They are rescaled at different scale factors using estimation based on the variable pixel spatial structure and common interpolation algorithms presents as optional tools of image editing software like Adobe Photoshop(R) (referred earlier as nearest neighbour, bilinear and bicubic algorithms).Finally it will be evaluate the quality of manipulated images and quantified the subjectivity of the analysis using the quadratic error along images pixel lines.
A. Bai(a)o
Escola Superior de Tecnologia do Institute Politécnico de Castelo Bronco, Avenida do Empresário 6000 Castelo Branco, Portugal
国际会议
The 12th Conference of the International Association for Mathematical Geology(第12届国际数学地质大会)
北京
英文
386-390
2007-08-26（万方平台首次上网日期，不代表论文的发表时间）