基于非线性模型的混合分形图像编码方法  

Hybrid Fractal Image Coding Method Based on Nonlinear Model

作　　者：王永军[1] 刘纪红[1] 熊飞文 徐孝仙 

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110004

出　　处：《控制与决策》2001年第B11期692-696,共5页Control and Decision

基　　金：辽宁省自然科学基金项目 (972 194)

摘　　要：提出一个非线性模型 ,将典型分形块编码的 8种对称变换推广到 6 4种 ,并通过定义特征值的方法对域块选择以及几何变换方式选择进行加速。基于区块邻域匹配的快速匹配策略 ,对匹配效果不佳的区块进行块截取编码 ,对平滑块采用均值和三级四叉树编码 ,使得综合压缩性能良好 。A nonlinear model is presented to generalize 8 kinds of symmetric transformations of classical fractal coding. The selection process of domain block and transformation way is speeded up by means of defining eigenvalues to 64 kinds. By adopting a rapid matching strategy based on matching in neighborhood of range block, non-matching blocks are coded by truncation coding, and smooth blocks are coded by average coding. Three-level quadtree coding scheme is used. The compression performance of this method is good, especially in coding time.

关 键 词：非线性模型 分形块编码 块截取编码 领域 图像编码 

分 类 号：TN919.81[电子电信—通信与信息系统]