Semi-automatic Removal of Foreground Stars from Images of Galaxies

Abstract

A new procedure, designed to remove foreground stars from galaxy profiles is presented. Although several programs exist for stellar and faint object photometry, none of them treat star removal from the images very carefully. I present my attempt to develop such a system, and briefly compare the performance of my software to one of the well known stellar photometry packages, DAOPhot. Major steps in my procedure are: (1) automatic construction of an empirical 2D point spread function from well separated stars that are situated off the galaxy; (2) automatic identification of those peaks that are likely to be foreground stars, scaling the PSF and removing these stars, and patching residuals (in the automatically determined smallest possible area where residuals are truly significant); and (3) cosmetic fix of remaining degradations in the image. The algorithm and software presented here is significantly better for automatic removal of foreground stars from images of galaxies than DAOPhot or similar packages, since: (a) the most suitable stars are selected automatically from the image for the PSF fit; (b) after star-removal an intelligent and automatic procedure removes any possible residuals; (c) unlimited number of images can be cleaned in one run without any user interaction whatsoever.