Help for CLEAN

PURPOSE:
To implement the "clean" algorithm, where a point spread function is
iteratively removed from an image in order to deconvolve it.

EXECUTION:
clean inp=(blurredimage,psf) out=(restored,residual,convolved) mode=...

where:
blurredimage   Is the image convolved with the point spread function.
psf            Is the point spread function.
restored       Is an image which when convolved by psf will be similar to 
               blurredimage. It will not resemble the scene before
               convolution however.
residual       Is the remaining data after many psf subtractions.
convolved      Is "restored" convolved by psf. Should closely resemble
               "blurredimage". 

NOTE: psf must be odd in size and the max pixel must be at the center.

METHOD:
see   http://www.astro.virginia.edu/~eww6n/math/CLEANAlgorithm.html

Clean computes an image which, when convolved by psf, produces the input
blurred image. It does this by iteratively subtracting the psf until
nothing is left, just noise. Clean can be used to study the structure of
small abjects, to permit psf changes, and to make maps of artifact
inconsistent with the psf (which cannot be modelled as superpositions
of psf's).

Note: The restored image will not resemble the true image before convolution
with the point spread function. It is merely a possible such image.
The rms nearest image is generated by program WIENER.
The most likely image is generated by program MEM.

HISTORY:
9-1-98  J Lorre. 
COGNIZANT PROGRAMMER:  Jean Lorre