基于方向性小波子树分形图像编码的改进算法  

An Improved Fractal Image Coding Algorithm Based on Oriented Wavelet Subtrees

作　　者：刘康[1] 王宣银[1] 

机构地区：[1]浙江大学流体传动及控制国家重点实验室,浙江杭州310027

出　　处：《电路与系统学报》2004年第5期25-29,共5页Journal of Circuits and Systems

基　　金：航天支撑技术基金资助项目(2001-HT-ZJDX)

摘　　要：本文在分析混合分形零树小波图像编码算法(FZW)优缺点的基础上,提出一种新的基于方向性小波子树的分形图像编码算法。该算法结合零树小波编码和分形编码,通过在匹配搜索过程中使用方向性 range 和 domain 子树,提高匹配精度,改善了传统分形小波图像压缩中的方块效应,更大限度的保留了图像的边缘信息。实验结果表明,该算法在提高压缩比和去除图像的方块效应方面,均取得了良好的效果。After analyzing the merit and disadvantage of hybrid fractal zerotree wavelet image coding algorithm, a new improved coding algorithm based on the IFS mapping in the DWT domain is proposed. In this algorithm, the performances as zerotree-based encoder and fractal image coder are combined in the wavelet domain. Using oriented range and domain subtrees in the matching process, the matching precision is enhanced. For this reason, the image can be reconstructed blocklessly with more edge information retained. Experiment results are satisfactory, especially the compression ratio can be improved and the block effect in the reconstructed image can be eliminated.

关 键 词：小波变换 分形图像编码 零树 图像压缩 方向性子树 

分 类 号：TN919.81[电子电信—通信与信息系统]