一种快速分形图像编码算法  

A speeding up fractal image coding method

作　　者：杨好庄[1] 赵耀[1] 

机构地区：[1]北京交通大学信息科学研究所,北京100044

出　　处：《信号处理》2008年第1期152-156,共5页Journal of Signal Processing

摘　　要：目前,分形图像编码技术最主要的缺点仍然是编码时间太长。因此,如何提高分形图像编码速度成为当前分形图像编码技术的研究热点。本文从分形图像编码通用公式推导出一个不等式,利用此不等式,可以预先排除大量不可能与值域块匹配的定义域块,从而减少值域块与定义域块的匹配计算,以此达到缩短编码时间的目的。实验结果表明,在解码图像质量基本不变情况下,本文的方法所使用的编码时间比Fisher方案所需的编码时间减少了很多。Expending much time in the fractal image encoding is still the main drawback of fractal image coding. At present, accelerateing fractal image coding is the hot research field. In this paper, a speeding up fractal image coding method based on a proposed inequality is proposed. By the use of this inequality, a lot of domain blocks are ejected before the matching computation between range block and domain block. The experiment results show that our method needs less encoding time and have the nearly same decoded image quality measured by peak-signal-to-noise-ratio compared with Fisher＇ s scheme.

关 键 词：分形图像编码 不等式 算法 值域块 

分 类 号：TN919.81[电子电信—通信与信息系统] TN919.8[电子电信—信息与通信工程]