快速分形图像编码的一种特征方法

快速分形图像编码的特征向量法是最具创新性、最有前途的方法之一,但它有几个缺点、特别是特征向量的高维数性.针对这个问题,本文提出减少分形编码时间的一种可选的特征方法.作为它的应用,本文先定义图像块的新特征——叉迹,然后提出一个基于叉迹的快速分形算法.这个算法把Range-Domain子块匹配问题转化为叉迹意义下的邻域搜索问题.对256×256 Lena图像的实验显示,与基于全搜索的基本分形算法比较,依赖于搜索邻域大小,该算法既能在峰值信噪比相同的情况下实现加快3倍多,也能在主观质量有一定下降的成本下实现加快100倍以上.

A Feature Method for Fast Fractal Image Encoding

Feature vector method for fast fractal image encoding is considered as one of the most innovative and promising approaches,but it suffers from several drawbacks,especially high dimensionality of feature vectors.Thus,an alternative feature method to reduce fractal encoding time is proposed.As one of its applications,cross trace-based fast fractal algorithm is presented,where the cross trace is a newly-defined feature of an image block.The proposed algorithm converts the range-domain block matching problem to the neighborhood search problem in the sense of cross trace.A simulation on popular 256×256 Lena image shows that,depending on the search window size,the proposed algorithm not only can achieve the speed-up of over 3 times with the same PSNR (peak signal-to-noise ratio) as the baseline fractal algorithm with the full search,but also can obtain the speed-up of 100 times or more at the cost of tolerable degradation of the decoded image quality.