基于相似比的快速分形编码算法  

A Fast Fractal Encoding Algorithm Based on Similar Ratio

作　　者：张爱华[1] 盛飞[1] 杨培[1] 常康康[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》2012年第11期176-178,共3页Computer Technology and Development

基　　金：国家自然科学基金(61070234);南京邮电大学校科研基金(NY210018)

摘　　要：对于分形图像编码时间过长这一问题,目前已经提出很多种解决方法,大多数都是基于分类或是特征匹配的搜索方法。把全局搜索转化为一定意义下的局部搜索是加速分形编码的一种有效可行方法。文中提出了一种基于相似比的分形编码算法并给出了可行性分析,该算法将全局搜索转化为相对意义下的近邻搜索,大大减少了搜索最佳匹配块的时间。本算法只搜索与值域块相似比相差较近的码本块,减少了搜索对象。实验结果表明,该算法在保证一定的重建图像质量前提下,大大提高了编码速度。To overcome the long encoding time of fractal image coding, many methods have been proposed and most of these are based on image blocks classification or feature matching. It is really a good choice to convert the entire search into the nearest neighbor search de fined under some situation. It proposes a fractal image encoding method based on similar ratio pwhich converts the entire search into the nearest neighbor search of similar ratio defined in this paper cutting plenty of time and also give the feasibility analysis. To reduce the searching space gthe searching target only includes the domain blocks which share the same similar ratio. By comparing with the basic fractal algorithm, the simulation result shows the proposed algorithm not only can gurantee the same image quality but also can improve the encoding speed.

关 键 词：分形 分形图像编码 相似比 图像块分类 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]