期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2008
卷号：XXXVII Part B2
页码：289-292
出版社：Copernicus Publications
摘要：Curvelet transformation proposed in 1999, on the basis of wavelet transformation, is a new multi-scale geometric analysis method. Besides scale parameter and location parameter, it adds an azimuth parameter in structure elements compared with wavelet transformation, which makes the Curvelet transformation express anisotropic singular boundary of lines or curves. So, Curvelet transformation is better than wavelet in geometric features of image, building its basis theory in image magnification. Inverting Curvelet transformation according to the new efficient, the magnified image can be gotten. By designing simulation, the original image is zoomed out to its 2,3,4,6 and 8 times, using cubic convolution, Curvelet transformation and bilinear interpolation. Compared to the original image, the results are showed as follows: whether from the qualitative or quantitative, its effect on image magnification is superior to the cubic convolution and bilinear interpolation
关键词：Curvelet Transformation; Image Magnification; measure result; average gradient
