自适应分裂－合并高压缩比图像编码

An Adaptive Split-and-Merge High Compression Image Coding

本文提出了一种自适应分裂－合并高压缩比图像编码方法，它将图像数据分解为与人眼视觉特性相匹配的分量，然后对各个分量分别进行编码，因此是一种高效的图像分割编码方法。为实现高压编比，本文还对控制图像的选择、纹理表示、分裂与合并应满足的条件和算法、边界编码和纹理编码等各个环节作了深入的分析，得出了一些有意义的结果。实验表明，在主观视觉质量仍保持较好的前提下，对一幅１２８×１２８的标准头肩像，压缩比可达２０：１左右。

A high Compression image coding method by means of adaptive split and merge algorithm is presented in this paper. In fact,the method partitions the original image data into different parts that match humar visual characterisites, and then each part effectively is encoded. Threfore,it is one of effective ways of image coding. To complete the coding procedure, this paper has also made mary research works on the selection of contral images,the representation of textures,conditions and algorithms that split and merge should satisfy, boundary coding and texture coding,etc.,and many meaningful results has got. The experiments show that,to an 128×12t image,The compression ratio could reach to 20:1.