deconvolution without PSF

[bob Woolery]

I am attempting to bring my homegrown image intensifier based X-ray system
into the ImageJ world.  I would love to deblur the images using one or
another deconvolution methods, but need to estimate the PSF.  Since the vast
majority of the resolution degradation is in the electron optics of the
image intensifier, a target of a small hole in a sheet of 5# lead seemed a
way to get an approximate PSF.  No dice, at least with the 1 mm target hole
I tried.  There seem to be blind or iterative methods, but I haven't found a
clearly applicable method.  It seems that a PSF should be derivable from
step function response, covering half the field with metal and exposing at
an X-ray flux that gives half or three quarters scale.  I recall from career
before last that PSF corresponds to 2D impulse response, and so the
derivation of PSF Could lead through Laplace Transforms, but I'm at sea here
and now.  The one reference I found was radio-astronomical and obscure to my
remaining mathematics.

 

bob Woolery, DC

stateoftheartchiro.com

326 deAnza dr

Vallejo, CA 94589

[phone]

 


--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html