Wavelet transform (CWT and DWT)

AI Thread Summary
Wavelet transforms, specifically Continuous Wavelet Transform (CWT) and Discrete Wavelet Transform (DWT), involve computing inner products between a signal and wavelets at various scales and translations. The DWT utilizes a bank of low-pass and high-pass filters to produce approximation and detail coefficients, which represent the signal's decomposition into wavelet and scaling functions. The coefficients are essential for reconstructing the signal and understanding its frequency content, with the spectrogram being limited by the wavelet's frequency range. The discussion also highlights the distinction between CWT and DWT, particularly regarding the absence of a father wavelet in CWT, which focuses solely on the mother wavelet's scaled and shifted versions. Understanding the relationship between these components is crucial for effective signal analysis using wavelets.
fog37
Messages
1,566
Reaction score
108
Hello,
I recently got interested in wavelets. The main idea seems clear: we compute the inner product between the signal ##x(t)## and a chosen wavelet for different scale factors and translations of the wavelet over the signal. The inner product provides the coefficient for a wavelet with a specific scale factor ##a##, which is inversely related to the wavelet frequency ##f##, as we translated the wavelet over ##x(t)##.

Apparently, given a discrete signal ##x(t)##, we can calculate either its continuous wavelet transform CWT and its discrete wavelet transform (DWT). Both are transforms are discrete in the sense that the scale parameter and translation parameter have a finite numbers of values...

My question: the DWT can be represented as a bank of low-pass and high-pass filters. We send the signal ##x(t)## into the first pair of filter and then pass its downsampled low-pass versions into subsequent filter pairs This process apparently produces approximation and detail coefficients...I am not clear on this process...What do we do with the approximation and detail coefficients? Is the signal decomposed into a weighted sum of wavelet functions plus a weighted sum of scaling functions?

We end up with a single downsampled low-pass version of the input signal and two downsampled high-pass versions....How does that relate to obtaining a spectrogram ##F(\omega, t)## of the input signal ##x(t)##?

1709516194114.png


Thank you!
 
Engineering news on Phys.org
If I'm not mistaken, the coefficients are displayed over a set band of frequencies (the wavelet). Whereas, Fourier displays the amplitude of the entire spectrum of frequencies. So, your spectrogram would be limited by the wavelet.
 
osilmag said:
If I'm not mistaken, the coefficients are displayed over a set band of frequencies (the wavelet). Whereas, Fourier displays the amplitude of the entire spectrum of frequencies. So, your spectrogram would be limited by the wavelet.
My understanding is that the detail and approximation coefficients will be the coefficients which will multiply the scaling (father) wavelets and the mother wavelets. The filters above are all bandpass filters with different frequency ranges....

I am not sure why the CWT, which is also discrete in the scale parameter ##a## and translation parameter ##\tau##, does not need the father wavelet....Any idea?
 
I guess I would say that once you have down or up sampled the signal, you have moved on to a different wavelet, with different a and t values. I would agree with you that it is a weighted sum of wavelet functions with their scalar coefficient.
 
osilmag said:
I guess I would say that once you have down or up sampled the signal, you have moved on to a different wavelet, with different a and t values. I would agree with you that it is a weighted sum of wavelet functions with their scalar coefficient.
Thank you!

And what is your intuition about the filters corresponding to scaled and shifted mother wavelets?
How do you factor in the father wavelet (scaling function) which is not present in the CWT that is only based on assembling the signal as a superposition of scaled and shifted mother wavelets (the daughter wavelets)?
 
I have recently moved into a new (rather ancient) house and had a few trips of my Residual Current breaker. I dug out my old Socket tester which tell me the three pins are correct. But then the Red warning light tells me my socket(s) fail the loop test. I never had this before but my last house had an overhead supply with no Earth from the company. The tester said "get this checked" and the man said the (high but not ridiculous) earth resistance was acceptable. I stuck a new copper earth...
Thread 'Electromagnet magnetic field issue'
Hi Guys We are a bunch a mechanical engineers trying to build a simple electromagnet. Our design is based on a very similar magnet. However, our version is about 10 times less magnetic and we are wondering why. Our coil has exactly same length, same number of layers and turns. What is possibly wrong? PIN and bracket are made of iron and are in electrical contact, exactly like the reference design. Any help will be appreciated. Thanks. edit: even same wire diameter and coil was wounded by a...
Thread 'Beauty of old electrical and measuring things, etc.'
Even as a kid, I saw beauty in old devices. That made me want to understand how they worked. I had lots of old things that I keep and now reviving. Old things need to work to see the beauty. Here's what I've done so far. Two views of the gadgets shelves and my small work space: Here's a close up look at the meters, gauges and other measuring things: This is what I think of as surface-mount electrical components and wiring. The components are very old and shows how...
Back
Top