[AstroPy] Past Mention of Source Extractor Command in IRAF?

Peter Erwin erwin at mpe.mpg.de
Wed Apr 29 00:58:41 EDT 2009

On Apr 29, 2009, at 2:25 AM, Wayne Watson wrote:

> Probably not, but is what you mentioned source extraction? The idea is
> that you take a picture of some part of the night sky, and then turn  
> it
> over to the source extractor. It twists, rescales, and turns your  
> image
> every which way until it matches 10 or more corresponding objects
> (stars, etc.) in a photographic atlas of some sort. If it looks right,
> you can then extract information about those objects and maybe others.
> It's an astrometric tool.

This sounds like a combination of two or three things: source  
catalog matching, and generating astrometric solutions.  I can't quite  
from your description how much of each you're interested in.  But here's
some background:

1. Source extraction: this refers to a program which analyzes a FITS  
image and
finds "sources" -- places that are brighter than the background, such  
as stars
and galaxies (but hopefully not random noise spikes).  It generates a  
table which
lists each source, its position, and various measurements derived
directly from the image (brightness, size, shape, orientation, etc.).   
It knows nothing
about other atlases or catalogs; it just reports what it can measure  
on that particular

Note that the position that the source extraction program reports are  
on the image in pixel coordinates (x, y).  If the FITS image happens  
to have
astrometric information in the header already (see below), then it  
will also calculate and
report the sky coordinates (RA, Declination) for each source.  The  
coordinates will be as good or bad as the astrometric information in  
the FITS
header.  A source extraction program does not do astrometry by itself  
at all; what
it does do is generate lists of sources which can then be used in the  

As others have pointed out, there are some tasks within IRAF that will  
do this.
There is also the standalone program SExtractor, which is designed to  
be run
from the command line (Unix/MacOS X,  or Cygwin).

2. Catalog matching: this means taking each source in the extracted  
catalog and
identifying it with a known astronomical object, such as a star in a  
star catalog.
Generally, this requires having good sky coordinates for the image.

If you *don't* have sky coordinates for the image, then you may need  
to match a few
objects (e.g. three or four bright stars) by hand (e.g., compare your  
image with a sky
atlas to figure out which star is which, then copy the RA,Dec  
coordinates by hand).  This
will give you a starting list for making an initial astometry  
solution. From there,
you can repeat the process (using the image with updated headers) and  
get a better
astrometry solution using the full list of sources from the source  
extraction stage.

3. Generating an astrometry solution: this means taking a matched list  
of object
coordinates in your image -- that is, a list of objects for which you  
have positions
on the image (pixel coordinates x, y) *and* their correct positions on  
the sky
(e.g., RA, Declination from a catalog) -- and then finding a general  
solution that
maps pixel coordinates (x,y) to sky coordinates (RA, Dec).  The  
resulting solution
is then stored in the FITS header, so that other programs can read it  
and convert
pixel coordinates (x,y) to sky coordinates (RA,Dec) automatically.

There are tasks within IRAF which do this; there is also a standalone  
called SCAMP (designed to work with catalogs produced by SExtractor).

The astrometry.net project is intended to be a sort of all-in-one  
solution for Step 3,
doing Steps 1 and 2 automatically along the way as needed.  If it  
works out as
planned (and it sounds like they're making progress), it will probably  
end up being
all you (or even most of us!) ever need.  I'm impressed as hell with  
their ambition.
But at the moment it's in an alpha stage; I've never tried it, the web  
interface requires
registering as an "alpha tester", and it looks like installing it on  
your own machine
would require several gigabytes of disk space.

In the meantime, you might also look at the programs here:



Peter Erwin                   Max-Planck-Insitute for Extraterrestrial
erwin at mpe.mpg.de              Physics, Giessenbachstrasse
tel. +49 (0)89 30000 3695     85748 Garching, Germany
fax  +49 (0)89 30000 3495     http://www.mpe.mpg.de/~erwin

More information about the AstroPy mailing list