北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2006, Vol. 29 ›› Issue (4): 33-35.doi: 10.13190/jbupt.200604.33.lihs

• 论文 • 上一篇    下一篇

新的学习矢量量化初始码书算法

黎洪松, 刘洪伟   

  1. 北京师范大学 信息科学与技术学院, 北京100875
  • 收稿日期:2005-04-01 修回日期:1900-01-01 出版日期:2006-08-30 发布日期:2006-08-30
  • 通讯作者: 黎洪松

A New Initial Codebook Algorithm of Learning Vector Quantization

LI Hong-song, LIU Hong-wei   

  1. College of Information Science and Technology, Beijing Normal University, Beijing 100875,China
  • Received:2005-04-01 Revised:1900-01-01 Online:2006-08-30 Published:2006-08-30
  • Contact: LI Hong-song

摘要:

针对原有随机数设置法、训练矢量集随机抽取法和LGB分裂法等初始码书算法存在的码矢利用率较低、运算量大和与信源匹配程度不高等不足,提出了一种新的分离平均法,并应用到基于自组织特征映射算法(SOM)的学习矢量量化(LVQ)中,图像矢量量化的实验表明,分离平均初始码书算法具有无效码矢数量少、码书性能高、运算量小、实现简单等优点。

关键词: 矢量量化, 初始码书, 自组织特征映射, 图像编码

Abstract:

In vector quantization(VQ),the initial codebook design is very important for VQ codebook performances. To overcome disadvantages of existing initial codebook algorithms, a new separating mean algorithm for learning vector quantization(LVQ)based upon self-organizing feature maps(SOM) was proposed. Experimental results for image VQ show that new initial codebook algorithm is better than random and splitting algorithm.

Key words: learning vector quantization, self-organizing feature maps, image coding

中图分类号: