Newsgroups: sci.image.processing
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news4.ner.bbnplanet.net!news3.near.net!paperboy.wellfleet.com!news-feed-1.peachnet.edu!news.Gsu.EDU!gatech!swrinde!cs.utexas.edu!news.sprintlink.net!EU.net!sun4nl!news.nic.surfnet.nl!frontier.tno.nl!usenet
From: Jos Groote Schaarsberg <schaarsberg@tpd.tno.nl>
Subject: Re: Resolution
Message-ID: <1995Jun26.174916.19831@frontier.tno.nl>
Sender: usenet@frontier.tno.nl (Usenet Postmaster id)
Nntp-Posting-Host: tpdfp1.tpd.tno.nl
Organization: TNO - The Netherlands Organization for Applied Scientific Research
References: <3sh35c$2f1e@usenetw1.news.prodigy.com> <perryDAs6w7.CsL@netcom.com>
Date: Mon, 26 Jun 1995 17:49:16 GMT
Lines: 52

>> Ron Tomlin (RKTR49A@prodigy.com) wrote:
>> Is there away to take a image, say from a frame grabber and zoom
>> in without losing the resolution ?
>>
> perry@netcom.com (Perry West) wrote:
>
> Once an image has been digitized in a frame grabber, there is a finite 
> amount of information present in the digitized image.  In this sense, 
> we are equating resolution with information content.  No amount of 
> additional processing can add information not present.  That is, you can 
> add useful resolution.  Zooming with interpolation may make the image 
> better for viewing for some purposes, but it does not add information.
> 

This answer (of three I read) corresponds the most with my ideas about
resolution. But a better description of "resolution in the view of
the application of Ron" would be helpfull. If the application is only
for e.g presentations, you can cheat and add all kinds of
(interpolated) information which looks nice for the human eye. 

I do not agree fully with the use of the word "information" as used
by Perry. Quite often with image processing, there is a model into
which we try to fit an acquired image. If this model is correct, we
can use characteristics of this model to extract non-acquired
information from an image.
Example 1: The worst image of a ring/circle has only to contain three
good parts to reconstruct the radius and position of this ring.
Example 2: We mostly use an acquisition system with the modulation
transfer decreasing at higher frequencies. If this MTF is well known,
the frequencies in a image can isolated and its amplitude be 
corrected with the inverse of the MTF.
Example 3: With some of our images, we know that the image of the
measured lightsource has a Gaussian profile with an offsets and skew
due to the e.g. dark current, Photo Response Non Uniformity, etc.
When the image has been defocused (enough but not to much) we can
perform calculations in order to determine the subpixel position of
the "centre of gravity".

With all these examples, we calculate parameters of the image
which corresponds to our expectations/model. These parameters can
be used to zoom-in on the model and the image.

If I guess, I expect that Ron needs a good model of his images and
their characteristics. With this model, he can construct the optimal
interpolation/extrapolation/... techniques to create a "sub-pixel"
resolution without adding wrong information.

Regards, Jos Groote Schaarsberg

P.S. I'm doing only the acquisition, not the processing so don't ask
for the algorithms of the examples!

