基于Krawtchouk矩不变量和核模糊聚类的快速分形编码

针对基于分类的快速分形编码方法存在着编码速度与解码质量间的矛盾,鉴于Krawtchouk矩不变量具有在仿射变换下保持不变的特性和核模糊聚类在处理非线性问题上的突出优势,本文首次将这两者引入到分形编码中,提出了基于Krawtchouk矩不变量和核模糊聚类的自适应分类快速分形编码方法。首先根据Domain块的方差将其粗分类,再根据Domain块的Krawtchouk矩不变量利用核模糊聚类对Domain块细分类。实验结果表明,与其他基于分类的快速分形编码方法相比,在解码图像质量提高的同时,大大加快了分形编码的速度。

Fast fractal coding method based on Krawtchouk moments and kernel-induced fuzzy clustering

There is the contradiction between coding speed and the quality of the decoded image in many fast fractal coding methods based on classification.In this paper,as the Krawtchouk moments are invariant under affine maps and the kernel-induced fuzzy clustering has prominent advantages in dealing with nonlinear problems,both of them are used to fractal coding,and a fast fractal coding method of adaptive classification based on Krawtchouk moments and kernel-induced fuzzy clustering is proposed.Firstly,the Domain ...