[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Lots of Data



"Creager, Robert S" wrote:
> 
> Why would Star (or any other software) need to change from the Mark III to
> the Mark IV images?  From what I understand, the Fits header contains all
> the image information necessary to go about analyzing it.  I know I'm
> missing something...

The Mark III camera is a "drift scan" camera.  The simple explaination
of "drift scan is that the camera is fixed. (does not track) but the 
digital "film" moves.  The Mark IV is a conventional camera.  The Mark IV
also has a much larger image size.

Yes, software could be written in a generalized way so that it could
handle both a Mark III and a Mark IV image.  IRAF (and I assume other
image processing systems) is that generalized but the program we used
for the Mark III, "star", is very specialized to the Make III and would
need a lot of work to handle Mark IV images.

> 
> I'll start working on a Perl script which will do the brief process Tom laid
> out.  I'll work on Linux, but the Perl scripts will be written to work on
> Windows also (hopefully).  Since IRAF is public domain with source, I'll
> probably spend some time getting it to run on Linux and figuring out how it
> works.

While far from being an expert, I have some experiance with IRAF.  First
off it may not be so easy to mix Perl and IRAF.  It can be done with
the latest version of IRAF now but you loose a lot.  IRAF has it's own
scripting language.  IRAF is also setup to process images in bulk.  For
example to apply a funtion to all images in a directory there is no
need to loop.  Just say "FunctionName *.FTS".  This is faster too
as the program "FunctionName" is only loaded once.

IRAF has both an Intractive mode and a scripted mode.  I find the way
you work with IRAF is at first by hand interactively.  Then once you
know how to do something you create a little script.  Later you may
want to string a few of these little scripts to gether.into a bigger
script.  You work your way up.  

Using IRAF here is what I'd do:

1) use Perl or shell script to dump all the images into the "correct"
directories for processing.  Change file names as required so they all
end in ".fits"

2) Patch up FITS headers to IRAF's liking.

3) Run "ccdproc" on the images.  This will do all the dark/flat
processing. trim the image section out, possibly remove cosmic rays
and baisically give you a bunch of cleaned up images.

4) run one of the photometric packages in IRAF against all the 
"cleaned" images.  There is both an Aperature and PSF fitting package.
We can debate which is best forever or run both and compare results.
The output of this step will be a star list for each image processed.

5) match each start list to a catalog to get good astrometry.  I think
the IRAF package to do this is called just "match".

6) Dump all the star lists into the site's database.

If you need some help getting IRAF up and working on your PC let
me know or post to the list and maybe some one else will help too.
Be warned, IRAF is the 600 pound gorilla of Image Processing packages.
I's big but don't let that stop you.  You will not need to used even
1/4 of it as much of it is for stuff like spectrography and other things
TASS does not do.
> 
> 1.  For each Fits image in a given location, do the following steps
> 2.  Parse Fits header to find sky location.
> 3.  Have IRAF (or other software) identify bright majors in image.
> 4.  Take stars found in 3 and compare against stellar catalog (presumably to
> find exactly where we are in the sky so lower magnitude stars can be
> matched?).
> 5.  Have IRAF (or other software) identify all (or to some magnitude) stars
> in image.
> 6.  Correct magnitude of all stars found (maybe from stars used in 4).
> 7.  Spit out a list of stars.
> 
> That about it for starters?  I know that eventually dark and flat frames
> will be needed, along with lots of other stuff, but it's a start...  I'll
> try to make the code easy to add new steps/change what software is used.
> 
> Later,
> Rob
> 
--
   Chris Albertson             home: chris@albertson-home.net
   Redondo Beach, California   work: calbertson@primeadvantage.com