作　　者： ; ; (张颖）;

机构地区： 广东省邮电管理局拓展办广州

出　　处： 《电子学报》 1998年第10期70-74,共5页

摘　　要： 待编码图像经过金字塔型离散小波变换后的系数在小波域内可以组成分层树状数据结构一小波树，这些跨越不同分辨率的小波树之间存在一定的相似性，可以通过分形变换来描述.本文正是在构造小波树的基础上，提出了基于分形和小波变换的自适应混合图像压缩算法.实验证明，我们提出的图像压缩方法与JPEG相比，能够在相近的压缩比的情况下（60：1～70：1），使得重建图像的PSNR（＞29.5dB）增加约5.4dB，并且图像的主观视觉质量也远远优于JPEG. Pyramidal wavelet decomposition provides a hierarchical data structure for image representation-wavelet subtree, which is composed of the coefficients at different resolution and different orientation but with same spatial location. In this paper, the conventional fractal coding in spatial domain is extended to wavelet domain by employing the self-similarities between different wavelet subtrees through proper affine transformation. Further, in order to achieve a good trade-off between image quality and compression ratio.an adaptive partition algorithm is given which can split the wavelet subtree into children subtrees according to local image complexity. Experimental results show that the new hybrid image compression algorithm can obtain much better coding performance in terms of PSNR and subjective quality as compared to JPEG at roughly the same compression ratio.

关 键 词： 分形编码 小波变换 图像压缩 混合编码

领　　域： [电子电信] [电子电信]