## Re: Correcting Old Blurred Photographs?

Expand Messages
• For those interested in pursuing a little of the more technical side of this subject, and who may be looking for actually spending some money on software to
Message 1 of 19 , Feb 28, 2004
For those interested in pursuing a little of the more technical side
of this subject, and who may be looking for actually spending some
money on software to do some of this (but almost all of it is rather
expensive for the uses most of us would have) go to the Reindeer
Graphics site below. It includes a good tutorial on various aspects of
image improvement. A couple of the links on that page appear to be
outdated, but if you use only the address up to the (For the record,
I have no connection to the company at all.)

http://www.reindeergraphics.com/tutorial/index.shtml

..............Mike

On Feb 28, 2004, at 3:59 AM, linsley@... wrote:

> It is actually quite possible. The math is rather hairy, but the
> process is fairly simple to program. The hard part is coming up
> with the right "point spread function." This is a mathematical
> description of what a single point looks like, after it is blurred.
> (Think of a blurred picture of a single star. The point spread
> function is a mathematical transformation that converts the ideal
> tiny dot into the blurred picture.) For a picture that is blurry
> because it is out of focus, it is simple; for motion blur, it is
> not. If the picture happens to contain a single (blurred) point,
> like a star, the function could be computed from the picture itself;
> otherwise, it would probably take a bunch of tweaking and trial-and-
> error to get a good picture. (Assuming, of course, you have the
> software to do it. Since it is fairly easy to program, I suspect
> someone has written it, but finding it is another matter. Learning
> to use it well enough to be worth the bother, is yet another.)
• Arghhhhhh.... I hate when that happens! I posted an incomplete edit of the message. Please ignore the line that starts with A couple of the links... ...
Message 2 of 19 , Feb 28, 2004
Arghhhhhh.... I hate when that happens! I posted an incomplete edit of
the message. Please ignore the line that starts with "A couple of the

:-|
.........Mike

On Feb 28, 2004, at 8:05 AM, Michael Stupinski wrote:

> For those interested in pursuing a little of the more technical side
> of this subject, and who may be looking for actually spending some
> money on software to do some of this (but almost all of it is rather
> expensive for the uses most of us would have) go to the Reindeer
> Graphics site below. It includes a good tutorial on various aspects
> of image improvement. A couple of the links on that page appear to be
> outdated, but if you use only the address up to the (For the
> record, I have no connection to the company at all.)
>
> http://www.reindeergraphics.com/tutorial/index.shtml
>
> ..............Mike
• ... Watkins, Sadun & Marenka. Modern Image Processing: Warping, Morphing and Classical Techniques. Academic Press Professional (a division of Harcourt Brace &
Message 3 of 19 , Mar 1, 2004
Susan Farmer <sfarmer@...> wrote:
> > I wrote:
> > > either way: In theory, it is probably possible, but this would be more
> > > akin to the sort of image processing that NASA does to enhance the detail
> > > in Hubble space telecope images, than it is to anything that consumer,
> > > or even pro, photo software can do.
> >
> > By a funny coincidence, today my employer was giving away some old
> > books, and I picked up one on image processing that happens to
> > discuss this very subject.
>
> Oooh. Reference? Book title?

Watkins, Sadun & Marenka. Modern Image Processing: Warping, Morphing
and Classical Techniques. Academic Press Professional (a division of
Harcourt Brace & Co.), 1993.

Note, however, that they do not really deal with correcting anything
other than the sort of generic, random (Gaussian) fuzzyness typical
of out-of-focus pictures or atmospheric turbulence in astronomical
photos. They mention that correcting other problems can be done if
you can find the point spread function, and that, if the image
contains a single point object (like a star), the point spread
function can be calculated from the image itself, but do not describe
how to do so, nor how to estimate the point spread function if the
image does not contain a point object.

--
Chuck Linsley
linsley@...
• Sorry for the delay, have been trying to find a scientific answer but haven t. All I know is that an old B&W pix looks a lot better under a dark blue light.
Message 4 of 19 , Mar 3, 2004
Sorry for the delay, have been trying to find a scientific answer but
haven't. All I know is that an old B&W pix looks a lot better under a
dark blue light.

Wilbur D. Russell
1015 South 4th
Leavenworth Ks.66048-3410
Pho 913/651-5662 QUE SERA SERA
http://community.webtv.net/wilburd/SLLUGseniorLansing
MAY THE SOURCE BE WITH YOU
• I have a Umax scanner here at school and it has a built-in descreen feature for scanning fine art prints (175 lpi), magazine pictures (133 lpi) and
Message 5 of 19 , Mar 25, 2004
I have a Umax scanner here at school and it has a built-in
descreen feature for scanning "fine art prints" (175 lpi),
"magazine pictures" (133 lpi) and "newsprint (85 lpi).

Is it possible to do this "descreen" procedure to an image
>after< it's been scanned? My scanner at home doesn't have
this feature and if I can get away with not having to rescan
a bunch of images, I'll be a happier camper. :-)

thanks!
Susan
Your message has been successfully submitted and would be delivered to recipients shortly.