图像矢量量化——改进的自组织特征映射神经网络的算法  

Improved Self-organizing Feature Mapping Algorithm for Vector Quantization of Images

作　　者：周文文[1] 董恩清[1] 

机构地区：[1]苏州大学电子信息学院,江苏苏州215021

出　　处：《通信技术》2009年第3期233-235,共3页Communications Technology

摘　　要：图像矢量量化(VQ)是图像压缩算法中的重要环节,在VQ中起决定性因素的是构造出性能优异的码书。为改善矢量量化码书的性能,文中在分析Kohonen自组织特征映射(SOFM)的基础上,提出一种识别距离SOFM的算法,同时将矢量量化应用于图像的小波变换域。测试结果表明,改进的算法使码书设计的计算量得到明显的降低,而且码书的性能得到了提高。Image Vector Quantization （VQ） plays an important role in image compression algorithm. A decisive factor in VQ is to construct the codebook of outstanding performance. In order to promote the codebook performance of vector quantization, this paper proposes an algorithm of Recognition Distance SOFM based on the basic self-organizing feature mapping （SOFM）. At the same time, the algorithm is applicable in wavelet transform of origin images. Simulation shows that the computation is substantially reduced and the codebook performance obviously improved.

关 键 词：自组织特征映射 神经网络 矢量量化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP183[自动化与计算机技术—计算机科学与技术]