New Image Matching Package - IMMATCH


A new package of image matching tasks is now available for user testing. The immatch package contains a set of routines for doing simple image astrometry, and for registering, matching the point spread functions, and matching the intensity scales of two or more 1D and 2D images using a variety of techniques. The current contents of the package are:
     ccmap - Compute image plate solutions using celestial coordinate lists 
    cctran - Transform from x-y to ra-dec and vice versa using plate solutions
  ccsetwcs - Create an image wcs from the plate solution
 ccxymatch - Match celestial and pixel coordinate lists using various methods
    geomap - Compute geometric transforms using matched coordinate lists
   geotran - Transform 1-D or 2-D images using the computed transforms
 geoxytran - Transform coordinate lists using the computed transforms 
   imctran - Transform from one image celestial wcs to another
  linmatch - Match the linear intensity scales of 1-D or 2-D images
linxymatch - Match pixel coordinate lists using various methods
  psfmatch - Match the point-spread functions of 1-D or 2-D images
    slalib - Starlink positional astronomy  library
  skyctran - Transform coordinate lists from one celestial wcs to another
    skymap - Compute geometric transforms using image celestial wcs information
skyxymatch - Generate matched x-y lists using image celestial wcs information
 sregister - Register 1-D or 2-D images using image celestial wcs information
   wcscopy - Copy the wcs of one image to that of another
  wcsctran - Transform image coordinates from one builtin image wcs to another 
    wcsmap - Compute geometric transforms using image wcs information
wcsxymatch - Generate matched x-y lists using image wcs information
 wregister - Register 1-D or 2-D images using image wcs information
 xregister - Register 1-D or 2-D images using x-correlation techniques

The ccmap, cctran, ccsetwcs, geoxytran, imctran, linmatch, linxymatch, psfmatch, skyctran, skymap, skyxymatch, sregister, wcscopy, wcsmap, wcsctran, wregister, and wcsxymatch tasks are new. Geomap and geotran are slightly modified versions of the corresponding tasks in the images packages. These modifications were necessary to properly support the new registration tasks. Xregister is already available under 2.10.3/2.10.4 and in the nmisc external package but was included here for completeness. Immatch is still under development and new tasks will be added in the future.

The major new capabilities of the package are ability to: 1) compute the wcs for an image given a matched list of pixel and celestial coordinates and store the wcs in the image header, 2) register images using accurate world coordinate information in the image headers, 3) match the image point-spread functions by computing the kernel required to convolve the higher resolution image to the resolution of the lower resolution image, 4) match the linear intensity scales of images using a variety of techniques including pixel-to-pixel comparisons and statistical measurements in one or more regions, and 5) perform celestial coordinate transformations on coordinate lists.

Some of this work was done in collaboration with Drew Phillips and is summarized in the article "Registering, PSF-Matching and Intensity-Matching Images in IRAF" (A.C. Phillips and L.E. Davis 1995, in Astronomical Data Analysis Software and Systems IV ASP Conference Series Vol. 77, eds. R.A. Shaw, H.E.Payne and J.J.E Hayes, p 297)

All the celestial coordinate transformations required by the ccmap, imctran, skyctran, skyxymatch, skymap, and sregister tasks are performed by routines in the Starlink positional astronomy library Slalib. Slalib has been made available to the IRAF project courtesy of Patrick Wallace of Starlink, and will become part of the iraf core system math package in the next major release of iraf.

These tasks are being made available prior to distribution as part of the IRAF 2.11 core system in order to invite user testing and comments. Please send any questions, comments, or problem reports to davis@noao.edu or iraf@noao.edu. The package retrieval and installation instructions are available in the readme file.


[IRAF logo] IRAF [NOAO logo] NOAO

Last updated: 17May1996