08-12-2013 01:58 PM
Thank you again for your quick reply please let me explain using mathematical equations as it will make more sense to you.
I have two functions (X1) and (X2), Then I take the FFT of both functions, fx1=fft(X1) and fx2=fft(X2)
Then I perform the cross correlation between both function which is : fX1X2 = fX1.*conj(fX2) ;
Then I perfom a normalization of the cross correlation
fx1x2phat = fx1x2./((abs(fx1x2)).^0.5); My question is: how do I do if I want to change the Value 0.5 from "0.1 to 0.9" with a step of 0.1
In Matlab the code would look at the following.
fx1=fft(x1);
fx2=fft(x2);
fx1x2 = fx1.*conj(fx2);
fx1x2phat = fx1x2./((abs(fx1x2)).^0.5);
A friend of mine also told me that I could directly put my Matlab code in LabView, but I do not know how to do it.
Thank you Israel
08-12-2013 02:01 PM
Thank you again for your quick reply please let me explain using mathematical equations as it will make more sense to you.
I have two functions (X1) and (X2), Then I take the FFT of both functions, fx1=fft(X1) and fx2=fft(X2)
Then I perform the cross correlation between both function which is : fX1X2 = fX1.*conj(fX2) ;
Then I perfom a normalization of the cross correlation
fx1x2phat = fx1x2./((abs(fx1x2)).^0.5); My question is: how do I do if I want to change the Value 0.5 from "0.1 to 0.9" with a step of 0.1
In Matlab the code would look at the following.
fx1=fft(x1);
fx2=fft(x2);
fx1x2 = fx1.*conj(fx2);
fx1x2phat = fx1x2./((abs(fx1x2)).^0.5);
A friend of mine also told me that I could directly put my Matlab code in LabView, but I do not know how to do it.
Thank you Israel
08-12-2013 02:30 PM
Don't overcomplicate things. LabVIEW has a function for crosscorrelation.
08-12-2013 03:48 PM
Thank you very much for your feedback but you have forgotten a very small and simple details.
The cross correlation you have presented to me is performed in the time domain, therefore require more storing resources than if it was performed in the frequency domain.
If someone can just explain to me how to use LabView to perform the computation presented in my previous email, I will be more than grateful.
To Be honest and fair, you guys are amazing.
Thank you
Israel,
08-12-2013 03:58 PM - edited 08-12-2013 04:09 PM
@israel.haifa wrote:
The cross correlation you have presented to me is performed in the time domain, therefore require more storing resources than if it was performed in the frequency domain.
Did you even read the help page I was linking above? The function is polymorphic and you can easily select frequency domain algorithms (it is even the default!!!).
08-12-2013 08:11 PM
Hello altenbach,
I would like first to start by thanking you for your reply, Why don’t you take the example giving to us by “Đặng Huy Hoàng”
Take two array of the same set of number: 1+1i , 2+2i, 3+3i, 4+4i then the conjugate will be 1-1i , 2-2i, 3-3i, 4-4i
Then do the cross correlation in the frequency domain using “Đặng Huy Hoàng” approach and your proposition and see if the results are the same. Please do not even worry about the normalization first.
Many thanks for your cooperation in this matter.
Israel