|Table of Contents|

A SOFM Algorithm for Vector Quantizing(PDF)

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

Issue:
1999年05期
Page:
393-396
Research Field:
Publishing date:

Info

Title:
A SOFM Algorithm for Vector Quantizing
Author(s):
Lin Chang Kang Taizhao ①
Fuzhou Telecom,Fuzhou 350005)
Keywords:
image processing neural netw orks feature ext ract ion
PACS:
-
DOI:
-
Abstract:
The characteristics of SOFM neural network is analysed and compared w ith the feature of Vector Quant izing problem in this paper. Based on this an algorithm for Vector Quantizing is put forward. Analysis and experiment s show that this alg orithm is stably convergent . The study result s are irrelevant to the initial status of the netw ork and quite approximate to the lower limit of the opt imum solutions. Very sat isfied results are achieved when this algorithm is used for g raphic data compression.

References:

1 Linde Y, Buzo A, Gray R. An algor ithm for vector quantizer design. IEEE Trans COMM, 1980,28( 1) : 84~ 95
2 Kohonen T , Barna G, Chr isley R. Statistical patter n recognition w ith neural networ ks: Benchmarking studies. ICNN- 88. California: IEEE San Diego Section and IEEE TAB Neural Networ k Committee, 19881I - 61~ I - 68
3 Pal N R, Bezdek J C, Tsao E C K. Gener alized cluster ing networks and kohonencs sel-f organizing scheme. IEEE T rans on Neur al Networ ks, 1993, 4( 4) : 549~ 557
4 Diamantini C, Spalvieri A. Certain facts about koho nenc s LVQ1 algorithm, IEEE Trans CAD,1996, 43( 5) : 425~ 427
5 杨行峻, 郑君里.人工神经网络.北京: 高等教育出版社, 1992.226~ 277
6 林昌, 康泰兆.基于自组织特征映射的栅阵列排序算法.电路与系统学报, 1997, 2( 1) : 40~43

Memo

Memo:
-
Last Update: 2013-03-29