首页 | 本学科首页   官方微博 | 高级检索  
     

神经网络-空间方向小波四叉树压缩编码
引用本文:任大伟. 神经网络-空间方向小波四叉树压缩编码[J]. 燕山大学学报, 2004, 28(4): 366-368
作者姓名:任大伟
作者单位:燕山大学,信息科学与工程学院,河北,秦皇岛,066004
摘    要:将空间方向小波四叉树编码与自组织特征映射神经网络相结合,提出了一种新的多尺度系数矢量量化压缩策略。首先通过小波分解得到三个方向的高频多尺度系数矢量,分别利用自组织特征映射神经网络对三个方向的多尺度系数矢量进行加权矢量量化压缩编码。仿真实验结果表明本文提出的算法是合理可行的。

关 键 词:多尺度  自组织特征映射神经网络  四叉树编码  小波分解  压缩编码  矢量量化  算法  高频  加权  仿真实验
文章编号:1007-791X(2004)04-0366-03
修稿时间:2004-01-08

Neural network-spatial orientation wavelet quad-tree image compression coding
REN Da-wei. College of Information Science and Engineering,Yanshan University,Qinhuangdao,Hebei ,China. Neural network-spatial orientation wavelet quad-tree image compression coding[J]. Journal of Yanshan University, 2004, 28(4): 366-368
Authors:REN Da-wei. College of Information Science  Engineering  Yanshan University  Qinhuangdao  Hebei   China
Affiliation:REN Da-wei1. College of Information Science and Engineering,Yanshan University,Qinhuangdao,Hebei 066004,China
Abstract:Based on spatial orientation wavelet quad-tree and self-organizing neural network, a new vector quantization image compression coding method is presented. Firstly three directional high frequent multi-scale coefficients vector is formed by wavelet decomposition, then self-organizing neural network is used to compress coding. Simulation results show that the algorithm presented is feasible.
Keywords:vector quantization  spatial orientation wavelet quad-tree  self-organizing neural network
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号