基于四叉树分割的分形图像编码的改进方法  

Improved fractal image coding based on quadtree partition

作　　者：吴红梅[1] 陈继荣[1] 

机构地区：[1]中国科学技术大学电子工程与信息科学系,安徽合肥230027

出　　处：《计算机应用》2005年第5期1090-1091,1095,共3页journal of Computer Applications

摘　　要：针对在分形图像压缩中,为寻找最佳匹配块需要花费相当长的时间的问题,在四叉树进行图像分割的基础上提出了一种基于分类函数和γ函数的分形图像编码方法。通过该方法,可以将图像子块分成24类,并有效地确定一种变换,而不需将每一定义域块进行各种变换,节省了搜索空间。实验结果证明,该方法可使编码速度有较大提高,而解码图像质量无明显下降。In fractal image compression, it takes quite long time to look for the best matching block.A method of fractal image coding was put forward based on quadtree image partition and class fuction as well as Y function. The method divides image blocks into 24 species, and then determines one transformation efficiently without carrying out various transformations on every definition domain block. Experiments prove that this method accelerates the speed of coding and the image quality of decoding dosen't decline obviously.

关 键 词：分形 四叉树 图像压缩 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]