FourierTransform produces valid results, but most users expect something else

[cesarsouza]

AForge.NET produces valid results in its FFT computation routines. However, over the years, some users have complained that either they do not produce the same results as MATLAB, that they have inverted signs, or were puzzled that they didn't work with non-square/power-of-two matrices:

http://code.google.com/p/aforge/issues/detail?id=100
http://code.google.com/p/aforge/issues/detail?id=156
http://code.google.com/p/aforge/issues/detail?id=221

As such, it would be nice to provide a version that addresses all those issues (even if the AForge.NET implementation is correct in the first place). The goal is to minimize user impedance while using the libraries.

[cesarsouza]

@andrewkirillov: @DiegoCatalano has contributed one new code to Accord.NET that seems to solve all those issues! However, even if I add it here, I am not sure if AForge.NET will have any future versions released... As such, I will include those fixes in Accord.NET itself and also port it to this fork. Please let me know what do you think... I can try to port it to the original AForge.NET repository as well!

Regards,
Cesar