一种吸引子图象编码的快速算法

A fast algorithm for attractor image coding

压缩速度慢是吸引子图象编码技术实用化的最大障碍 ,因此 ,如何加速其编码过程 ,是当今吸引子图象编码的研究热点。本文提出一种快速吸引子编码的方法 ,它将分形维数作为衡量值域子块和定义域子块是否匹配的主要依据 ,有效避免了传统编码方法中子块匹配的耗时运算 ,同时采用气泡上浮法和二分法等快速排序和查找方法加速其编码过程。实验结果表明 ,该方案比传统 Jacquin方法快

Slow compression speed is the greatest barrier for the practical use of attractor image coding, so how to speed up its compression procedure is a research hotspot now. In this paper, a fast algorithm for attractor image coding is proposed which uses fractal dimension to evaluate the matching degree of range blocks and domain blocks and therefore the time exhausting matching process of the tranditional fractal coding is greatly reduced. Meanwhile, bubble method and dichotomy are used to speed up the sorting and searching processes. Experimental results show that the speed of the scheme proposed in the paper is about 5 times as fast as traditional Jacquin's scheme.