Re: De-screening

[Kennedy McEwen <rkm@xxxxxxxxxxxxxxxxxxxx>]

In article <NEBBKHCFDGIGIJEHDJFCIEMJDEAA.pderocco@ix.netcom.com>, Paul 
D. DeRocco <pderocco@ix.netcom.com> writes
>For the scientifically inclined: The optimum algorithm for removing a 
>fixed-pitch halftone screen, I believe, would be to do two-dimensional 
>windowed Fourier transforms on overlapping blocks of the image, and 
>then notch out the spectral peaks that are caused by the screen, 
>without removing the higher-frequency energy. I doubt any inexpensive 
>scanner actually does this--my Epson 1640 certainly doesn't.

And for the historically inclined: I can remember doing exactly this 
process in analogue optics in my physics undergraduate course - 
initially using bits of sticky tape to mask out the peaks in the FT 
image before optically re-transforming it back to get a de-screened 
image.  :-)

Perkin-Elmer made a very nice "optical correlator" in the 1980s (which 
was publicly exhibited - so I am not divulging state secrets!) using 
this same approach, to automatically recognise and position targets in 
real world images for target tracking systems in missiles.  Faster than 
even today's supercomputers.  ;-)
-- 
Kennedy
Yes, Socrates himself is particularly missed;
A lovely little thinker, but a bugger when he's pissed.
Python Philosophers
-
Turn off HTML mail features. Keep quoted material short. Use accurate
subject lines. http://www.leben.com/lists for list instructions.