We discuss the derivations and manipulation of image source catalogs. Our initial catalogs may contain only X,Y positions with a few simple moments-based image parameters. More advanced catalogs will contail star-galaxy classification and more precsise photometric parameters.

1. The problem: What discrete sources are present in the image.
2. Making initial catalogs.
3. Reviewing initial catalogs.
4. Preparation for refined photometry (photom2).
5. Some simple aperture photometry.

A catalog of the dicrete sources present in an image is essential to nearly all factes of image reduction. Initially we may desire a simple list of X,Y positions for use in deriving the image world coordinated system (WCS). In later phases of the analusis we may require more precise photometric parameters for each image in order to perform source classification or derive a photometric zeropoint. This document was initially part of a specialized description of how I reduc HET acm (aquisition camear) images.

For many reasons, the first thing we must do is establish a WCS for the FIT image header, and for this, we need a list of the X,Y positions of the discreete sources on the image. Since this procedure is usually more time-consuming than many interactive processes, we desire a means of running this in a bact mode. The method below enables this. It is used for batch processing of catalogs in reduction reduction scripst like AcmRunDate.

 
 
Reduction location:  /home/sco/ACM_work_Oct2019
A good example that has a good number of stars, and a good WCS solution is: 
   /home/sco/ACM_work_Oct2019/red_20191022/local_red/FIXUP/20191022T025524.2_acm_sci.fits
   /home/sco/ACM_work_Oct2019/red_20191022/local_red/WCS/20191022T025524.2_acm_sci.fits

First I create a test directory: /home/sco/ACM_work_Oct2019/Test_imcats
% cat list.1
/home/sco/ACM_work_Oct2019/red_20191022/local_red/FIXUP/20191022T025524.2_acm_sci.fits

To make the catalog:
%  image_catalogs list.1 none N N  
Usage: image_catalogs list.0 none N N 
arg1 - Name of file with list of input FITS images
arg2 - output directory name appendix (can be "none")  
arg3 - Clean all but the table files (Y/N)   
arg4 - run in debug mode (Y/N)

Here is what results from the above run: 
% ls ./local_red/IMAGE_CATS/20191022T025524.2_acm_sci/
20191022T025524.2_acm_sci_BGboxes.reg  20191022T025524.2_acm_sci.fits	      20191022T025524.2_acm_sci.props
20191022T025524.2_acm_sci_bkg1.fits    20191022T025524.2_acm_sci_label0.fits  20191022T025524.2_acm_sci_sigma.fits
20191022T025524.2_acm_sci.cat0	       20191022T025524.2_acm_sci.params       20191022T025524.2_acm_sci.table
20191022T025524.2_acm_sci_detsig.fits  20191022T025524.2_acm_sci.parlab
 

It is often useful to make a quick visual check of the image catalogs compiled in the previous section. There are tools for this, but the user must run them from the same working directory where the catalogs were made (i.e. you must be at ./local_red/). Note that we can delete all the files in this working directory if we desire, but we must run our review from there.

 
 
% ls
local_red/  S/

To get a regions file (using the X,Y centroids):
% imgcat0_ds9reg_make ./local_red/FIXUP/20191022T025524.2_acm_sci.fits XY cyan N  

Usage: imgcat0_ds9reg_make ./local_red/FIXUP/20191022T025524.2_acm_sci.fits XY red N  
arg1 = name of FITS image run through image_catalogs 
arg2 = Use XY or Ra,DEC for coordinates (XY,RD) 
arg3 = color to be used fro region markers  
arg4 = run in debug mode? (Y/N)  

% ls
20191022T025524.2_acm_sci.cdfp	20191022T025524.2_acm_sci.cdfp.reg  local_red/	S/

% head -5 20191022T025524.2_acm_sci.cdfp
Ra_hrs,Dec_deg,Xcent,Ycent,isomag30,NumPix,
 0.00 	 0.00 	  737.84	  209.03	  12.182	  2081
 0.00 	 0.00 	  472.52	  90.99	  14.993	  468
 0.00 	 0.00 	  168.15	  756.53	  15.865	  365
 0.00 	 0.00 	  562.44	  186.81	  16.122	  320
% head -5 20191022T025524.2_acm_sci.cdfp.reg
image
circle(737.84,209.03,5.00) # color=cyan width=4 font="helvetica 20 normal roman" text={1}
circle(472.52,90.99,5.00) # color=cyan width=4 font="helvetica 20 normal roman" text={2}
circle(168.15,756.53,5.00) # color=cyan width=4 font="helvetica 20 normal roman" text={3}
circle(562.44,186.81,5.00) # color=cyan width=4 font="helvetica 20 normal roman" text={4}

NOTE that I could use the ./local_red/FIXUP/20191022T025524.2_acm_sci.fits name. This file 
actually does not exist  at this location, but only the basename is extracted for use in 
imgcat0_ds9reg_make. The important files do exist in ./local_red/IMAGE_CATS/20191022T025524.2_acm_sci/. 

Finally I have the imgcat0_ds9reg_make call made and used in images_catalogs_review/, and I 
can see the overplotted regions file in the process of reviwing the catalogs:  

If a version of our image with a valied WCS is available, then we can use that image to make a more informative cdfp file, namely one with accurate Ra,Dec values.

 
 
% best_wcs_image ./local_red/FIXUP/20191022T025524.2_acm_sci.fits N  
% cat best_wcs_image.out
./local_red/WCS/20191022T025524.2_acm_sci.fits
% cat best_wcs_image.btype
wcs

First, we need a file named "Plate.Scale" that will be used to record the plate scale 
% image_ps ./local_red/WCS/20191022T025524.2_acm_sci.fits > Plate.Scale
% cat Plate.Scale
0.27094460 0.00000173

We could make a more informative cdfp file with: 
% imgcat0_ds9reg_make ./local_red/WCS/20191022T025524.2_acm_sci.fits RD green N  

% head -5 20191022T025524.2_acm_sci.cdfp
Ra_hrs,Dec_deg,Xcent,Ycent,isomag30,NumPix,
  19.740600	  26.221525000	  737.84	  209.03	  12.182	  2081
  19.739346	  26.207641667	  472.52	  90.99	  14.993	  468
  19.736712	  26.249802778	  168.15	  756.53	  15.865	  365
  19.739687	  26.216397222	  562.44	  186.81	  16.122	  320
%  head -5 20191022T025524.2_acm_sci.cdfp.reg
fk5
circle(296.10900879,26.22152519,6.97") # color=green width=4 font="helvetica 20 normal roman" text={1}
circle(296.09017944,26.20764160,3.31") # color=green width=4 font="helvetica 20 normal roman" text={2}
circle(296.05065918,26.24980354,2.92") # color=green width=4 font="helvetica 20 normal roman" text={3}
circle(296.09530640,26.21639633,2.73") # color=green width=4 font="helvetica 20 normal roman" text={4}

The final routine to review our list of catalogs is rather simple to execute.

 
 
% image_catalogs_review list.1
**** Bingo, this works! All of the more that 300 detections are overplotted correctly in 
     in frame 3. 

It seems have a whole section on a prep code, but as is oftern the case, the bookwork involved in compileing a code can be more complicated and as critical as the actual code purpose. The code we want ti run, in this case photom2, will compute refined photometric parameters for the moments-based image parameters in the imgcat0 catalogs. Hence, we need acces to the imgcat0 information, specficially the X,Y location of the sources and the boundary box limits established for each source by the 4-connected pixel algorithm used in the imgcat0 routine.

The preparation code I developed for this is photom2_prep.

My first impulse was to develop photom2 not quite from scratch, but nearly so. I have a large number of well-tested image photometry routines in SCOLIB (namely SCOLIB.m2048ims.f) that power routines like ds9_imstats and mido.sh. Actaully the ds9_imstats code is just a wrapper that allows the user to build a ds9 regions file. The image processing work is done by misdo.sh, which measures photometric parameters for each aperture. A good first step would be experiment with these codes to derive aperture photometry at the imgcat0 p X,Y positions. This was deemed prudent before I went hacking up well-used library codes! The code developed for this is apphot.