I'm trying to find the rotation center between two images taken at different position angles. The trivial case of 180 degrees is simple enough, but how would I find the rotation center for an arbitrary relative rotation theta?

I've tried using geomap with fitgeom='rotate' but I am not getting consistent results. I am at the point where I can easily create a matched catalog of (x1,y1) (x2,y2) coordinates for the same source in the two images. Any ideas? Am I missing something obvious?

Dave (LBT Observatory)

Hi Dave,

I am not aware of a task or existing recipe to find the rotation center. I'm not sure if GEOMAP can do what you want. If you figure something out feel free to post it. Other than getting a reply from other readers of this forum I think you are on your own.

Yours,
Frank Valdes

I am at the point where I can easily create a matched catalog of (x1,y1) (x2,y2) coordinates for the same source in the two images. Any ideas?

If by this you mean you have a list of coordinate files for a reference image and an "unknown" image, then see XYXYMATCH or more specifically the help page examples. You can also use e.g. WREGISTER to register the images according to the WCS so that the coordinates match up better. Post back if I've misunderstood your question.

Cheers,
-Mike

No, I have a single file with x1, y1, x2, y2 data as input for geomap (example below). Each line has the coords for the same object in two images rotated by some amount (several hundred matches). The two images were taken with only a rotation between them, so ultimately there should be a common RA,DEC on the same X,Y coords in the two images.

550.01 1298.25 1504.00 4593.45
545.22 1306.97 1508.86 4584.37
365.89 1307.92 1688.00 4583.41
192.86 1318.30 1860.70 4572.69

XYXYMATCH and WREGISTER do not seem to determine what I need. Both, like GEOMAP, determine rotation and offset from arbitrary centers.

I am trying to determine the XY coords of the center of rotation for the camera rotator. This is where the pointing model origin needs to be for good open-loop tracking and rotation.

That you thinking about this:
http://iraf.net/phpBB2/viewtopic.php?t=85811&start=0&postdays=0&postorder=asc&highlight=

No, I already know the rotation angles between the two images. I am trying to find the rotation center.

I posted a reply to your problem of relative rotations under that topic.