机构地区: 浙江大学理学院计算机图象图形研究所
出 处: 《软件学报》 2003年第4期864-868,共5页
摘 要: 在分析彩色图像色彩三分量r,g,b的相关性和分形四叉树编码层次信息冗余性的基础上,提出了一种分形彩色图像压缩编码方法.它将图像的3个独立的颜色分量按某种方式组合成1个来搜索匹配块,从而将需要存储和搜索的3个颜色分量匹配块(SFC方法)减少为1个,并且对四叉树层次信息进行压缩.此外,采用不同的组合,得到了几个图像压缩比和解码质量相近的编码方法,其中使用亮度分量的方法比使用其他方法速度更快.实验结果表明,它优于SFC方法及标准JPEG方法,不失为一种好的分形彩色图像压缩方法. In this paper, a color image fractal compression coding method is proposed by analyzing the correlation of r, g, b components of a color image and the remnant of layer information in quadtree scheme. It merges three-color components into one. Accordingly this method will reduce the matching blocks, which need to be saved and encoded by SFC (separated fractal coding) algorithm, into one and at one time the source data and layer information in quadtree scheme can sustain non-losses and high-proportion compression. Furthermore, by selecting different combinations, a series of coding methods that bear very close compression ratio and quality of decoded image are obtained. Moreover the coding speed of the luminosity method among them is faster than the other ones. Experimental results indicate that this approach excels SFC or standard JPEG and can be a good color image fractal compression algorithm.