Estimation of global and local holder exponent (values) directly from wavelet transform modulus maxima(WTMM)

2 visualizzazioni (ultimi 30 giorni)
Dear Friends
I am very new to signal processing techniques . Urgently need to do some Wavelet analysis on time series. I used matlab "wavemenu" graphical user interface for continuous wavelet 1-D and got 3 variables(coefs,scales,wname), But dont know how to generate WTMM (wavelet transform modoules maxima ) , how to export the values of WTMM , and more important, how to estimate global and local holder exponent (values) directly from WTMM representation?
I did some experiment using Wavelab850 toolbox but got stuck over it since it requires many input parameters to be set
Pls do help me if you have any idea.
Regard

Risposta accettata

Wayne King
Wayne King il 18 Ago 2011
I'd have to think a bit about how to actually estimate the Holder (Lipschitz) exponents from the CWT, but you can use localmax in the Wavelet Toolbox to extract the local maxima lines.
For example:
load cuspamax;
x = 1:length(cuspamax);
scales = 1:32;
cfs = cwt(cuspamax,scales,'haar');
[lmaxima,indices] = localmax(cfs,[],false);
[iRow,iCol] = find(lmaxima);
subplot(211);
imagesc(abs(cfs)); axis xy;
axis([1 1024 1 32]);
ylabel('Scale'); title('CWT Coefficients (Moduli)');
subplot(212);
plot(x(iCol),scales(iRow),'marker','o','markerfacecolor',[0 0 1],...
'linestyle','none');
xlabel('Position'); ylabel('Scale'); title('Maxima Lines');
axis([1 1024 1 32]);

Più risposte (1)

Changxi You
Changxi You il 9 Mag 2016
We can use localmax to get the lines of maxima. Anybody knows how to estimate the holder exponents after that?

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by