基于小波与分形理论的图像压缩编码算法  

Image Compression Coding Algorithm Based on Wavelet and Fractal Theory

作　　者：张爱华[1] 何雨虹[1] 张璟[1] 

机构地区：[1]南京邮电大学理学院,江苏南京210046

出　　处：《计算机技术与发展》2017年第6期46-50,共5页Computer Technology and Development

基　　金：国家自然科学基金面上项目(11471114;61372125);江苏省自然科学基金项目(BK20150867);南京邮电大学攀登计划一项(NY210018)

摘　　要：在分形图像编码中,影响分形图像编解码速度的主要因素是从大量码本中搜索R码本块的最佳匹配码本块。如果能够使用一种方式尽可能缩短匹配块的搜索范围,那么编码的时间就可以大大减少。然而,在提高编解码速度的同时,重构图像的质量却有所降低。针对上述这些问题,在定义一种图像子块的新特征—欧氏比基础上,将小波变换与分形编码有机结合,提出了一种基于小波与分形理论的图像压缩编码算法。该算法将全局搜索码本块转化为局部搜索码本块,缩短了编解码的时间,同时利用连续小波变换的平滑特性,进一步提高了重构图像质量。仿真实验结果表明,与特征算法中的梯度算法相比,所提出的算法不仅缩短了图像编解码的时间,还提高了重构图像的质量。In fractal image coding,it is the main factor that affects the decoding speed to search the best D sub block of the R sub block from a large number of code-book. However, the time of encode-decode would be significantly reduced if the search range of blocks could be cut down as much as possible. It is the problem that while the speed of encoding being improved the quality of reconstructed im- age would be getting worse. Aiming at above problems, after a new feature of the image sub-blocks, Euclidean ratio, has been defined, a fractal coding algorithm integrated wavelet transform with fractal encoding has been proposed, which converts the global search code- book to the local search code-book for having shortened the decoding time and employs the smooth property of continuous wavelet trans- form for improvement of the reconstructed image quality. The simulation results show that the proposed algorithm has not only decreased encode-decode time but also promoted the reconstructed image quality compared with the relative gradients algorithm of the characteristic algorithm.

关 键 词：分形图像编码 小波变换 欧氏比 子块特征 

分 类 号：TN919.81[电子电信—通信与信息系统]