Machine Vision

cancel
Showing results for 
Search instead for 
Did you mean: 

display FFT of a signal

Solved!
Go to solution

Hello

 

I have a noisy signal to process (Please find enclosed an image of the signal) and I would like to use the FFT function on my signal to filter the noise and retrieve the clean signal by using the FFT inverse.

I have trouble when I use FFT because I just have one peak while I am expected to have a curve and the the noise on higher frequencies.

 

Thanks

 

0 Kudos
Message 1 of 10
(4,858 Views)

You didn't attach the data

Randall Pursley
0 Kudos
Message 2 of 10
(4,823 Views)

Here is the signal I want to process

 

Thanks

 

0 Kudos
Message 3 of 10
(4,817 Views)

I'm sure there is other signals present, but your one peak is at DC (0 Hz) and is orders of magnitude larger than any other frequency present.  If your signal really starts at 28000, at least subtract that out from all points before performing the FFT.  You might try viewing it on a log scale as well.

Randall Pursley
0 Kudos
Message 4 of 10
(4,811 Views)

Hello

 

Please find enclosed the output of my signal using FFT. I have complex values in my output array.

How can i extract the noise now to clean the signal and retrieve it in the temporal field.

 

Thanks for your help

0 Kudos
Message 5 of 10
(4,801 Views)

Can you post your data instead of just a picture of your data?

Randall Pursley
0 Kudos
Message 6 of 10
(4,788 Views)

Here is my VI with the image

 

Let me know if it is enough

Thanks

Download All
0 Kudos
Message 7 of 10
(4,776 Views)
Solution
Accepted by topic author luckyloic

Take a look at the attached and see what I added.  Changes the window size will change the cutoff for the filtering.  I offset the result by one just to make it easier to see in the Waveform Graph.

Randall Pursley
0 Kudos
Message 8 of 10
(4,646 Views)

Hi

Thank you for the attached file.

I have some questions to understand your VI

What is the meaning of the window? Does it represent how much noise is extracted. What represents the center of hte window where the peak is located.

 

Else it works very well, thanks.

0 Kudos
Message 9 of 10
(4,627 Views)

The window is a lowpass filter.  This is a basic square window.  The center represents 0 Hz (or DC).  Positive frequencies to the right, negative frequencies to the left.  So this windows zeros out the frequency components that fall outside the window.

Randall Pursley
0 Kudos
Message 10 of 10
(4,611 Views)