DSP 版 (精华区)
发信人: grey (nemo*statics&posibility), 信区: DSP
标 题: 小波变换
发信站: 哈工大紫丁香 (2003年11月07日14:20:58 星期五), 站内信件
发信人: blizzard (蓝色困惑), 信区: signal
标 题: 小波变换
发信站: 武汉白云黄鹤站 (Tue Feb 29 14:27:01 2000), 站内信件
小波变换就是wavelets transform和小波分析wavelet analysis,
在数字信号变换和处理的领域得到比较广泛的应用。这里连载几片
文章对一些小波分析的基本理论和网络资源提供一些信息。
首先来看看小波的基本理论:
我们可以来看这样一个频域,分成L=+U-(其中+=(0,∞),-=(-∞,0)。
那么在L的2次方(也就是(0,∞))处我们引入一个窗口函数P(ξ),而且
有∑k∞=-∞P2(2-kξ)=1和P(ξ)在(3/4,3)处成立。这样我们可以清楚的
看到P(2-kξ)在(2k,2k+1)间隔内是一个窗函数,同时我们来看下面的函数:
Sk,j=Sin[(j+1/2)л(ξ-2k/2k)]P(2-kξ)
和Ck,j=Cos[(j+1/2)л(ξ-2k/2k)]P(2-kξ)
如果我们定义Sk,j为一个奇函数在sk,j范围内,而Ck,j为一个偶函数
在ck,j范围内,那么很明显可以看到Sk,j和Ck',j'是正交的。我们可以
写成Ck,j土iSk,j=e土ijлξ/2kφ(ξ/2j)在φ(ξ)=ejл/2ξP(ξ)处
是小波函数的fourier变换。
Thus, wavelet analysis corresponds to windowing frequency space in ``octave'
' windows
(2k,2k+1).
A natural extension is provided by allowing all dyadic windows in frequency
space and adapted window choice. This sort of analysis is equivalent to wave
let packet analysis.
The actual fast wavelet packet analysis algorithms (wavelets being a special
cases) permit us to perform an adapted Fourier windowing directly in time d
omain by successive filtering of a function into different regions in freque
ncy. The dual version of the window selection provides an adapted subband co
ding algorithm.
The wavelet packet library is constructed by iterating the wavelet algorithm
. This library contains the wavelet basis, Walsh functions, and smooth versi
ons of Walsh functions called wavelet packets (cf. [5])
Figure A.2: Wavelet-packets
These waveforms are mutually orthogonal. Moreover, each of them is orthogona
l to all of its integer translates and dyadic rescaled versions. The full co
llection of these wavelet packets (including translates and rescaled version
s) provides us with a library of ``templates'' or ``notes'' which are matche
d ``efficiently'' to signals for analysis and synthesis (cf. [2]), Wavelet p
acket expansions correspond algorithmically to subband coding schemes and ar
e numerically as fast as the FFT.
--
I'm Sailing; I'm Flying
Through The Dark Night; Far Away
I'm Dying; Forever Crying
To Be With You; To Be Free!
※ 来源:.武汉白云黄鹤站 bbs.whnet.edu.cn.[FROM: 203.95.7.153]
(本文采用S-Term文章拷贝脚本拷贝)
==================================================
--
※ 来源:·哈工大紫丁香 bbs.hit.edu.cn·[FROM: 202.118.250.183]
Powered by KBS BBS 2.0 (http://dev.kcn.cn)
页面执行时间:3.444毫秒