On Mon, Jun 15, 2009 at 03:59:57PM +0200, Gregor Hlawacek wrote:
> On Mon, 2009-06-15 at 14:57 +0200, David Necas (Yeti) wrote:
> > 
> > Now we are getting somewhere...  So what you actually need, is to
> > calculate PDSF not just along x averaged in y[*] or vice versa, but also
> > in other directions, right?
> 
> Yes that would be nice too. But I assume due to the averaging this will
> not give the same result as doing a cross section (without or minimal
> averaging) in the FFT.

Cross-section through the origin in the Fourier space is exactly the
same as averaging in the orthogonal coordinate for FT coefficients (this
can be proved trivially) and approximately the same (i.e. with a random
phase assumption) for their norms, i.e. for the PSDF.

Which means I might calculate it as 2D FFT modulus section in practice
anyway.  Even though this means somewhat more noisy result (`speckles')
than with averaging in the orthogonal coordinate.

> Something like a rotating line section coming from the center of the
> 2D-FFT were one can select the desired direction interactively would be
> cool. Like the current cross section tool but with a fixed origin of the
> line in 0,0.

I can imagine that.  And now someone needs to write the code...

Regards,

Yeti


------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables unlimited
royalty-free distribution of the report engine for externally facing 
server and web deployment.
http://p.sf.net/sfu/businessobjects
_______________________________________________
Gwyddion-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/gwyddion-users

Reply via email to