🚀 go-pugleaf

In article 1@ntuvax.ntu.ac.sg, egntoh@ntuvax.ntu.ac.sg () writes:
>I would like to hv information on image restoration in the area of
>motion blurr and out_of_focus restoration. Could anyone comment on the
>availability of good papers and software? By the way. is there any ftp
>site providing softare or images on image restoration ?

I am interested in this myself, however I would like to use motion blur
and defocus as a measure of image quality.  ANy papers on this subject?
In particular I am trying to automatically measure the quality of scanned
documents.  My Initial feeling is to look at spectral composition of 
portions of the text using a DCT...blurred images should be different than
non-blurred images, but this may be difficult to detect in frequency
domain.  Any ideas?

Thanks,

Lew Berman

egntoh@ntuvax.ntu.ac.sg writes:

>I would like to hv information on image restoration in the area of
>motion blurr and out_of_focus restoration. Could anyone comment on the
>availability of good papers and software? By the way. is there any ftp
>site providing softare or images on image restoration ?

I have written a Khoros Toolbox which contains several image restoration
routines. Before you can use it, however, you need to obtain Khoros.
Khoros is available through anon ftp at pprg.eece.unm.edu.
See also the newsgroup comp.soft-sys.khoros

-Mathew

>In article 1@ntuvax.ntu.ac.sg, egntoh@ntuvax.ntu.ac.sg () writes:
>>I would like to hv information on image restoration in the area of
>>motion blurr and out_of_focus restoration. Could anyone comment on the
>>availability of good papers and software? By the way. is there any ftp
>>site providing softare or images on image restoration ?
>


imdisp for MSDOS systems is good at removing blur or motion distortion.

I generally use these commands to do it.

Command: enh baa 30
Command: enh med

The med enhancement will put some blur back into it, I do that to remove the 
jaggedness that the baalke enhancement creates.


Op jck


-- 
Colby Kraybill
Space and Planetary Image Facility
University of New Mexico

In article <1992Jul24.105328.1@ntuvax.ntu.ac.sg> egntoh@ntuvax.ntu.ac.sg writes:

> I would like to hv information on image restoration in the area of
> motion blurr and out_of_focus restoration. Could anyone comment on the
> availability of good papers and software? By the way. is there any ftp
> site providing softare or images on image restoration ?

The normal linear way to do this is via the cepstral filtering: linear
motion blur results in two sharp dips in the cepstrum, while defocus
blur results in the superposition of a Bessel onto the cepstrum (which
you can identify by the zero crossings).

Probably the standard papers in this area are:

@ARTICLE{Cannon,
   AUTHOR = "T. Michael Cannon",
   TITLE = "Blind Deconvolution with Spatially Invariant Image Blurs with
            Phase",
   JOURNAL = "IEEE Transactions on Acoustics, Speech, and Signal Processing",
   VOLUME = 24, NUMBER =1, PAGES = "58--63",
   YEAR = "1976",
   KEYWORDS = "enhancement, got"}

@ARTICLE{Cannon-Trussell-Hunt,
   AUTHOR = "T. Michael Cannon and H. Joel Trussell and Barry R. Hunt",
   TITLE = "A Comparison of Image Restoration Methods",
   JOURNAL = "Applied Optics", VOLUME = 17, PAGES = "3384--3390",
   YEAR = "1978",
   KEYWORDS = "enhancement, got"}

Note that cepstral filtering will result in ringing at sharp edges
(after all, you will be throwing away the high frequencies).  To do
any better than this, you have to go to non-linear techniques such as
maximum entropy (MEM) or maximum _a posteriori_ (MAP).

Hope this helps.

--
 Dr Adrian F. Clark                                   JANET: alien@uk.ac.essex
 INTERNET: alien%uk.ac.essex@nsfnet-relay.ac.uk          FAX: (+44) 206-872900
 BITNET: alien%uk.ac.essex@ac.uk              PHONE: (+44) 206-872432 (direct)
 Dept ESE, University of Essex, Wivenhoe Park, Colchester, Essex, C04 3SQ, UK.

In article <ALIEN.92Jul28092012@vulcan.essex.ac.uk> alien@essex.ac.uk ("Adrian F Clark") writes:

>Note that cepstral filtering will result in ringing at sharp edges
>(after all, you will be throwing away the high frequencies).  To do
>any better than this, you have to go to non-linear techniques such as
>maximum entropy (MEM) or maximum _a posteriori_ (MAP).

Not necessarily.  See

@article{lagendijk88,
    TITLE = "Regularized Iterative Image Restoration with Ringing Reduction",
    AUTHOR = "R.  L.  Lagendijk and J.  Biemond and D.  E.  Boekee",
    JOURNAL = "IEEE Transactions on Acoustics, Speech, and Signal Processing",
    YEAR = "1988",
    VOLUME = "36",
    NUMBER = "12",
    PAGES = "1874-1888",
    MONTH = "Dec."}
 
for a method that uses space-variant linear techniques to avoid ringing.
It's quite powerful and works well.


--
Stan Reeves
Auburn University, Department of Electrical Engineering, Auburn, AL  36849
INTERNET: sjreeves@eng.auburn.edu