基于HVS分类及邻域搜索的快速分形图像编码

针对基本分形图像编码算法时间过长的问题，提出了一种基于HVS分类及邻域搜索的快速算法。根据HVS特性将子块分为平滑类子块和非平滑类子块，对于平滑类子块直接存储其均值，以减少需要搜索匹配父块的子块数；对于非平滑类子块，从离其最近的父块开始搜索，在搜索父块时，剔除与当前子块的近似度不满足要求的父块，并引入误差阂值和搜索父块的最大次数来控制子块的搜索过程。实验结果证明，该算法大大提高了编码速度。

Fast Fractal Image Encoding Based on Classification and Neighbor-Searching

In order to solve the problem of time consuming in the encoding process of the basic fractal algorithm, a fast method based on classification and neighbor-searching is proposed. Every range block is classified into smooth or non-smooth block according to the feature of HVS （Human Visual System）. The mean will be kept for every smooth range Mock, so that the number of range block which needs to search for the best domain block is reduced. Every non-smooth range block searches for the best domain block in its neighborhoods. The domain block which doesn＇t accord with kick-out condition based on the similar measure is excluded from the codebook. Furthermore, an error threshold and the most times of searching domain block are used to control the searching range. The experimental results demonstrate that the proposed algorithm is much faster than the basic fractal algorithm.