摘要
提出一种图像插值算法,将图像像素看成矩形块而非像素点,利用插值图像的矩形像素块由尺寸放大(或缩小)的原始图像的像素所占的百分比,将原始图像做放大或缩小变换,从而在最大程度上保留了原始图像的数据.即使经过多次的缩放变换,也不至使图像失真.
Taking the pixels of an image as rectangular area rather than point sets, an image interpolation algorithm is proposed, in this paper, by employing the idea that the original image is expanded (or contracted) by determining the percentage of rectangular pixels of the size expanded (or shrunk) for old image overlapping the interpolated images pixels. This method can retain much more of the original data than any other algorithm.
出处
《广西师范学院学报（自然科学版）》
2003年第4期72-76,共5页
Journal of Guangxi Teachers Education University(Natural Science Edition)
关键词
图像处理
插值
缩放变换
image processing
interpolation
expansion and contraction transformation
