ATM Removing image distortions

[Fabio Pettinati <fabio@voreas.com>]

On Thursday, January 23, 2003, at 03:52  AM, Vladimir Galogaza wrote:

> it is well defined. So the question is how to calibrate deformations.
> Can it be done by photographing well known dense star fields and 
> compare
> measured coordinates on the film with known coordinates of imaged 
> stars...


Take a look at this site:
http://home.arcor-online.de/axel.mellinger/mwpan_web/mwpan_web.html

It pretty much provides all the information needed to address the 
problem.

Please note that instead of photographing star fields, one could 
photograph a target of known dimensions at a sufficiently long distance 
and accomplish essentially the same result in terms of deriving the 
mathematical transformations. The target could be as simple as a sheet 
of paper where you print horizontal and vertical lines at the same 
distance (i.e. a square grid) using a laser printer. You could 
photograph that sheet during the day, or you could print a negative 
image (i.e. white grid on a black background--toner is cheap) and 
photograph it at night with proper illumination (sideways).

If the target is sufficiently far away the distortion seen on the 
photograph is essentially the same as when you photograph something at 
infinity. The closer the target to the camera (or telescope) the more 
different the distortion.

Fabio
San Jose, CA