Post Reply 
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Determining camera+lens distortions with Scamp
11-03-2016, 15:39 (This post was last modified: 11-03-2016 15:40 by Kimmo.)
Post: #1
Determining camera+lens distortions with Scamp
Hi

I am digitizing old astronomical photographic glass plates with a digital camera, and I want to determine the distortions caused by the camera+lens combination.
I have got the following recipe:

1) Plot an artificial star field on a paper (just small dots on paper)
2) Take images of the field at different orientations
3) Use Sextractor to find the stars and their positions
4) Use the obtained 'star' catalogues as input to Scamp, and setting the following parameters in Scamp:

MATCH N # Do pattern-matching (Y/N) ?
SOLVE_ASTROM Y # Compute astrometric solution (Y/N)
ASTREF_CATALOG NONE

Furthermore, I need to set
CROSSID_RADIUS 3000.0 # Cross-id initial radius
or otherwise I get
WARNING: Not enough matched detections in instrument A1

Scamp gives the following information about the catalogs:

Group 1: 5 fields at 4.37e+03 2.93e+03 with radius 5259
instruments epoch center coordinates radius scale
distort-1.cat A1 P1 +4.3675e+03 +2.9275e+03 5259. 1.000
distort-2.cat A1 P1 +4.3675e+03 +2.9275e+03 5259. 1.000

and the same for all the rest of the catalogs.
Obviously, all the values are in pixel units.

However, Scamp crashes in the following way:

> Making mosaic adjustments...
> Solving the global astrometry matrix...
> WARNING: Not a positive definite matrix in astrometry solver

----- Astrometric clipping:
> Computing astrometric stats for group 1
> out of memory, core-dump

Does anybody has a solution for this ?

cheers, Kimmo
Find all posts by this user
Quote this message in a reply
11-18-2016, 10:54
Post: #2
RE: Determining camera+lens distortions with Scamp
Hi

I have a solution for this.

I need to make an artificial image AND a corresponding star catalog.
So the catalog is like any other catalog that Scamp can use, having the positions of 'stars' in RA/Dec coordinates.
Then I just run Scamp for every image of my artificial stellar field, getting the corresponding '.head' file for every image. The '.head' files look like the following:

.
.
CTYPE1 = 'RA---TAN' / WCS projection type for this axis
CTYPE2 = 'DEC--TAN' / WCS projection type for this axis
CUNIT1 = 'deg ' / Axis unit
CUNIT2 = 'deg ' / Axis unit
CRVAL1 = 9.210773210248E-04 / World coordinate on this axis
CRVAL2 = 1.113289610851E-04 / World coordinate on this axis
CRPIX1 = 4.300000000000E+03 / Reference pixel on this axis
CRPIX2 = 2.900000000000E+03 / Reference pixel on this axis
CD1_1 = -2.442920305476E-06 / Linear projection matrix
CD1_2 = 9.543673356062E-09 / Linear projection matrix
CD2_1 = 9.543673356062E-09 / Linear projection matrix
CD2_2 = 2.442920305476E-06 / Linear projection matrix
PV1_0 = -3.149556166061E-05 / Projection distortion parameter
PV1_1 = 1.001363443070E+00 / Projection distortion parameter
.
.

In all the .head files the CRVAL* and CD* keywords are different.

I want to determine the distortions caused by camera+lens as an average over several of these images. So the question is, how to make an average of the 'Projection distortion parameters' (PV parameters) ? Is the average of PV1_0, for example, an average over the PV1_0 parameters in all the .head files ?

cheers, Kimmo
Find all posts by this user
Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)