Whitening the signal and finding the peak spectral power

Can anyone expalin me this "To identify the fundamental and harmonic frequencies of the contaminating movement with peak spectral power, the movement signal is spectrally whitened and the peaks were identified. This whitening is done to automate the detection of the spectral peaks with improved accuracy. The reference signal is then bandpass filtered to each of the identified frequencies ()"

1)how can Ispectrally whiten the signal

2) how can I find the spectral peaks of the signal

3)How can I find the frequqncy related to these peaks

[Mathieu NOE]

hello

I think yu have to "remove" the contaminating peaks (that you can identify in the spectrum) by extrating from the signal the bandpass filtered version of signal to recover the "clean" (whitened) signal.

so basically you have to :

1/ identify the peaks (frequency, width)

2/ create the bandpass filters and generate filtered signals (one per filter)

3/ remove from the signal all filtered signals (you should no have a clean signal with "white" (flat) spectrum)

[Mathieu NOE]

hello again

you may also consider this :

SPECTRAL WHITENING - File Exchange - MATLAB Central (mathworks.com)

[Celestie Gladys]

yes i did somthing like this !

Answer 1

[Akash]

Hi,

I understand that you are seeking assistance with spectral whitening, identifying spectral peaks, and finding the corresponding frequencies in MATLAB.

To spectrally whiten a signal, you can use the MATLAB function "Spectral whitening" available on the MATLAB File Exchange. The link is provided below. This function generates a flat Fourier spectrum for a non-white signal, either across the full frequency range or within a user-defined frequency band.

https://www.mathworks.com/matlabcentral/fileexchange/65345-spectral-whitening

The output "xnew" represents the spectrally whitened signal, which spans the entire frequency range from 0 Hz to the "Nyquist frequency".

To find spectral peaks in a signal, you can utilize the "findpeaks" function in MATLAB. This function identifies local maxima in the signal, which correspond to the peaks. To access the documentation for the "findpeaks" function, you can follow the link provided below.

https://www.mathworks.com/help/signal/ref/findpeaks.html?s_tid=doc_ta

To determine the locations of the peaks in the signal, you can refer to the "Find Peaks and Their Locations" section within the documentation of the "findpeaks" function.

If you encounter difficulties in determining the frequencies associated with the peaks, you can refer to the below link which offers insights and solutions on how to find specific frequencies in a signal.

https://www.mathworks.com/matlabcentral/answers/475727-how-to-find-specific-frequency-in-signal?s_tid=answers_rc1-1_p1_Topic