RE: [APML] Giant Scorpius Tail Montage

["Chris Schur" <cschur@therim.com>]

Ray, the distortion of schmidt images is legendary around the periphery and
matching them is an art.  Even my best software cannot do some negs.  This
one was close but despite weeks of work, is not perfect.  If youve ever
tried doing a panorama with regular lenses, you can imagine the distortion
around the edge!  Mine is not lens caused, but by the curvature of the film
holder thats akin to a basketball.

Chris

-----Original Message-----
From: astro-photo-bounces@seds.org
[mailto:astro-photo-bounces@seds.org]On Behalf Of Ray Butler
Sent: Monday, August 15, 2005 2:06 AM
To: Discussion of Film Astrophotography
Subject: Re: [APML] Giant Scorpius Tail Montage



Chris,

That's an excellent mosaic. I don't see any seams resulting from
intensity scaling issues, which is impresssive! However, along the
junction between the top and middle panels, there is a curious geometric
distortion - everything looks slightly stretched. Since I don't see the
same thing on the other junction, it can't be related to the schmidt
camera optics - I think it must be in the processing.

This distortion reminds me of what I used get if I screwed up a
high-order geometric transformation (registration) between two images in
IRAF. E.g. if one of the tie-points I entered - supposedly the same star
on the 2 images - was actually 2 completely different stars, due to a
mixup on my part. The software does its best to fit the transformation,
but since it is based on flawed positional input, it ends up causing
distortion like that. I used have to go back through the list of
tie-points to track down my star mismatch. (Later on, I realised that I
could ask to software to use automatic statistical rejection of dodgy
tiepoints...which weeded out any error on my part! (;0)).

I wonder if you had a similar problem with your mosaicing software? Did
you have to tell it which stars were in common on the two images, or
does it do some automatic pattern recognition to register the images?
Even in the latter case, things can go wrong...

Ray "who believes that humans still beat computers at visual pattern
recognition - but also that we are a tad more prone to mental fatigue"
Butler

--
Dr. Ray Butler
Lecturer, Physics Department & Computational Astrophysics Laboratory,
National University of Ireland - Galway,
University Road, Galway, Ireland.
Web: www.nuigalway.ie/physics/ 	Email: ray.butler-AT-nuigalway.ie
Tel: +353-91-493788 		FAX: +353-91-494584

_______________________________________________
Astro-Photo mailing list
Astro-Photo@seds.org
http://seds.org/mailman/listinfo/astro-photo
---



---

_______________________________________________
Astro-Photo mailing list
Astro-Photo@seds.org
http://seds.org/mailman/listinfo/astro-photo