1

The wavelet transform informs what all derived wavelets (from mother wavelets) with different scales at different point of time , are present (along with the associated strength) in my original signal. I am not able to understand how do I now correlate it with FFT that simply gives what all sine and cosine waves are present.

Now how do I make use of this information that I get from WT? I cannot say unlike as in case of FFT or STFT " we have a 100hz frequency in the window of 4 seconds". I will have to say "we have a db4 wavelet present in some window". How it will make sense. How do I interpret this in a meaningful way?

gpuguy
  • 1,370
  • 8
  • 17
  • 32
  • http://dsp.stackexchange.com/questions/651/which-time-frequency-coefficients-does-the-wavelet-transform-compute – endolith Jun 19 '13 at 20:48

1 Answers1

2

Wavelets are used for time-scale, or time-frequency, analysis, just like STFT. You can map the coefficients of CWT to frequencies through the center frequency of the wavelet. See for example http://www.mathworks.se/help/wavelet/ref/scal2frq.html

jiihaa
  • 21
  • 1