icwtft
This function is no longer recommended. Use icwt
instead.
Syntax
xrec = icwtft(cwtstruct)
xrec = icwtft(cwtstruct,'plot')
xrec = icwtft(cwtstruct,'signal',SIG,'plot')
Description
cfs field of the
structure array cwtstruct. Obtain the structure
array cwtstruct as the output of cwtft.
xrec = icwtft(cwtstruct,'plot')
plots the reconstructed signal.
xrec = icwtft(cwtstruct,'signal',SIG,'plot')
places a check box in the bottom left corner of the plot. Enabling the
check box superimposes the plot of the input signal
SIG on the plot of the reconstructed
signal. By default the check box is not enabled and only the
reconstructed signal is plotted.
Input ArgumentscwtstructStructure array containing six fields.
cfs — CWT
coefficient matrix
scales — Vector of
scales
frequencies —
frequencies in cycles per unit time (or space)
corresponding to the scales. If the sampling
period units are seconds, the frequencies are in
hertz. The elements of frequencies are in
decreasing order to correspond to the elements in
the scales vector.
omega — Angular
frequencies used in the Fourier transform
meanSig — Mean of
the analyzed signal
dt — The sampling
period
wav — Analyzing
wavelet used in the CWT with parameters
specified
cwtstruct is
the output of cwtft.
Output ArgumentsxrecReconstructed signal
Examples
Compute the CWT and inverse CWT of two sinusoids with disjoint
support.
N = 1024;
t = linspace(0,1,N);
y = sin(2*pi*8*t).*(t<=0.5)+sin(2*pi*16*t).*(t>0.5);
dt = 0.05;
s0 = 2*dt;
ds = 0.4875;
NbSc = 20;
wname = 'morl';
sig = {y,dt};
sca = {s0,ds,NbSc};
wave = {wname,[]};
cwtsig = cwtft(sig,'scales',sca,'wavelet',wave);
% Compute inverse CWT and plot reconstructed signal with original
sigrec = icwtft(cwtsig,'signal',sig,'plot');
Select the check box in the bottom left corner of the
plot.
Use the inverse CWT to approximate a trend in a time series.
Construct a time series consisting of a polynomial trend, a
sinewave (oscillatory component), and additive white
Gaussian noise. Obtain the CWT of the input signal and use
the inverse CWT based on only the coarsest scales to
reconstruct an approximation to the trend. To obtain an
accurate approximation based on select scales use the
default power of two spacing for the scales in the
continuous wavelet transform. See cwtft for details.
t = linspace(0,1,1e3);
% Polynomial trend
x = t.^3-t.^2;
% Periodic term
x1 = 0.25*cos(2*pi*250*t);
% Reset random number generator for reproducible results
rng default
y = x+x1+0.1*randn(size(t));
% Obtain CWT of input time series
cwty = cwtft({y,0.001},'wavelet',{'bump',[4 0.7]});
% Zero out all but the nine coarsest scale CWT coefficients
cwty.cfs(1:80,:) = 0;
% Reconstruct a signal approximation based on the coarsest scales
xrec = icwtft(cwty);
plot(t,y,'k'); hold on;
xlabel('Seconds'); ylabel('Amplitude');
plot(t,x,'b','linewidth',2);
plot(t,xrec,'r','linewidth',2);
legend('Original Signal','Polynomial Trend',...
'Inverse CWT Approximation');
figure
plot(t,x,'b'); hold on;
xlabel('Seconds'); ylabel('Amplitude');
plot(t,xrec,'r','linewidth',2);
legend('Polynomial Trend','Inverse CWT Approximation');
You can also use the following syntax to plot the approximation.
Select the box to view the original polynomial trend
superimposed on the wavelet approximation.
% Input the polynomial trend as the value of 'signal'
xrec = icwtft(cwty,'signal',x,'plot');
More About
Inverse CWT
icwtft computes the inverse CWT based on a
discretized version of the single integral formula due to
Morlet. The Wavelet Toolbox™ Getting Started Guide contains a brief
description of the theoretical foundation for the single
integral formula in Inverse Continuous Wavelet Transform. The
discretized version of this integral is presented in [5]
References
[1] Daubechies, I. Ten Lectures on
Wavelets, Philadelphia, PA: Society
for Industrial and Applied Mathematics (SIAM),
1992.
[2] Farge, M. “Wavelet Transforms and Their
Application to Turbulence”, Ann. Rev.
Fluid. Mech., 1992, 24,
395–457.
[3] Mallat, S. A Wavelet Tour of Signal
Processing, San Diego, CA: Academic
Press, 1998.
[4] Sun,W. “Convergence of Morlet's Reconstruction
Formula”, preprint,
2010.
[5] Torrence, C. and G.P. Compo “A
Practical Guide to Wavelet Analysis”,
Bull. Am. Meteorol. Soc.,
79, 61–78, 1998.
Introduced in R2011a