近邻相关匹配图象快速分形压缩编码

A RAPID FRACTAL COMPR ESSION CODING FOR THE IMAGE NEIGHBORHOOD CORRELATION MATCHING

为缩短分形编码中值域块与定义域块逐一匹配时间 ,将值域块与定义域块的全局匹配改为近邻匹配。并提出近邻相关匹配分形近似编码的方法 ,在子块搜索匹配之前将大量显著不相关的子块剔除掉 ,从而大大缩小搜索空间。通过实验 ,该方法比全局搜索匹配的分形编码快。

I n order to shorten the one-by-one match time between the range block and the domain block in fractal coding,the global matching between that is changed into the neighbor matching and the neighbor fractal approximation coding based on correlation matching is proposed.Before the sub-block is searched and matched,a large number of obviously uncorrelative blocks are deleted.So the searching space is quretly narrowed.Viatesting,this mathod is faster than fractal coding based on the global searching matching.