一种新的用于LBG图像压缩初始码书生成算法

LBG算法是一种经典的矢量量化算法,广泛用于数字图像的压缩.在矢量量化中,码书的设计问题实际上解决的是如何选择最具有代表性的矢量作为码字来尽可能精确表示整个矢量空间.码书决定了矢量量化的编码质量,因此高质量的码书设计就成为整个矢量量化研究的核心.初始码书的质量极大的影响了LBG算法的收敛速度和收敛的可能性.并决定了最终的码书质量.针对常用的算法生成的码书分布不能很好的代表样本集的分布,并造成LBG算法收敛缓慢和最终码书的品质不佳,提出一种新的快速的初始码书生成算法,提高了生成码书的迭代速度和最终码书的品质.

A new algorithm research of primary codebook for image compression

LBG, widely used in digital image compression, is a classical vector quantization algorithm. In the vector quantization codebook designing is to select the best vector (codeword) that can represent the entire vector space as precise as possible. The codebook decides the coding quality of vector quantization. Therefore the high quality codebook design becomes the core of research for entire vector quantization. A new algorithm of primary codebook for image compression is presented in this paper. This approach provides a practical solution to eliminate LBG algorithm codebook and empty cyst problems. The simulation results show that the new algorithm increases speed of codebook production and reduces time of computing. A good codebook can be obtained.