IMAQ Correlate shows strange result

[comrade]

Hi,

when using the IMAQ Correlate function from IMAQ 7.0 with 8-bit grayscale pictures I receive a result picture which looks a lot like a FFT picture. Though I am not too familiar with what the correct result should look like I am quite sure that a 2D-symmetric picture is not what should be there. I would expect something like high grayscale values, where the correlation is high and low values where it is low....

I have tried to cope with this issue using the inverse FFT function from IMAQ, but then I either get errors from the Correlate function or the FFT function, since invFFT uses Complex images, while Correlate uses only 8-bit. Also when converting 8-bit image to an array, converting this to CSG-type array and then converting back to a complex image I still get errors for the CSG array to Complex Image function.

Can anyone give me hint on this one?

[ThSa]

Comrade,
there is another forumpost about this issue:
http://forums.ni.com/ni/board/message?board.id=200&message.id=3881&requireLogin=False

If you post your application incl. example pictures I will have a look on it...

Regards,
ThSa

[comrade]

ThSa,

thanks for the hint on the other thread in the froums - together witha student co-worker we resolved th "problem". I had a false perception of what the correlation should have as a result. After some tests with easy images we found that the IMAQ Correlate does just what it is supposed to. The thought that the picture shows an FFT-image turned out to be completely wrong, although it looked very much like it.

The questioned remains, however, why converting the image the following way:
8-bit --> 2D-array of U8 --> 2D-array of CSG --> complex image
still yields an error for the CSG-->complex image conversion.  Currently I am writing from home, where I don't have the programming example available. I will return to office some time next week where I can drop the files, if you are interested...

Thanks.
Falk aka Comrade