基于二维特征的快速分形图像编码方案

Fast fractal image coding scheme based on two-dimension feature

分形图像编码方法因其具有高压缩比、重建速度快、分辨率无关性等优点使得它有很好的发展前景.但是,编码过程需要在一个海量码本中寻找每个range块的最佳匹配domain块而耗去特别多的时间,这阻碍了它的广泛应用.为了加快编码过程,提取图像快的能量和方差来构成它的二维特征,作为图像子块间的相似性度量,在range块与domain块匹配时,大多数与range块的二维特征不相近的domain块,可以在匹配前排除.因此,对一个range块,能够在较小的搜索范围内找到它的最佳匹配domain块,使编码时间极大地减小.仿真实验显示,它能够大大缩短编码时间,同时实现和全搜索分形编码方法相近的重建图像质量.

The fractal image encoding method has much better potential for its many advantages,such as high compression ratios,fast decoding and resolution independence.However,it has not been widely applied due to exhaustive inherent encoding time,which is mostly spent on searching for the best-matched block to every range in a usually large domain pool.To speed up fractal encoding process,the energy and variance of an image block form its the two-dimension feature that is used as the criterion for the similarity of two image blocks,the most inappropriate domain blocks,whose two-dimension feature are not similar to that of the given range block,are excluded before matching during the domain-range matching process.Therefore,for a range block being encoded,its best-matched domain block can be captured in the less search space,so the encoding time is reduced greatly.Simulation results demonstrate that,for three standard test images,the proposed scheme significantly shortened the encoding time,while providing the close image quality as full exhaustive search fractal coding method.