Thread: noise level comparison in Matlab

Hello;

Is there a good way to compare noise levels between noisy and de-noised 1-D signals in Matlab? Something that would actually show a visible difference... The noise I used was generated by rand. Fast Fourier Transform...? Power Spectral Density? I'm not sure where to start. I'm cleaning signals with wavelets and I want some proof to show that the noise level has been reduced other than just looking at the de-noised signal. Suggestions will be greatly appreciated. Thanks.

Originally Posted by disclaimer

Hello;

Is there a good way to compare noise levels between noisy and de-noised 1-D signals in Matlab? Something that would actually show a visible difference... The noise I used was generated by rand. Fast Fourier Transform...? Power Spectral Density? I'm not sure where to start. I'm cleaning signals with wavelets and I want some proof to show that the noise level has been reduced other than just looking at the de-noised signal. Suggestions will be greatly appreciated. Thanks.

PSD would be a good start.

If you know the signal you started with, comparison of de-noised signal with the ground truth signal in both time and frequency domain might be nice.


CB