6 Removing image contamination

 6.1 Removing consistent image faults
 6.2 Removing other image contamination

It is usually the case that any image containing a galaxy of interest will also contain a number of other objects such as galaxies, stars, cosmic rays or even image flaws. Those that fall on the image far beyond the apparent limits of the galaxy may be ignored, but those lying close to the galaxy of interest must be removed. The is because the profiling applications have been written under the assumption that the only sources of pixel count on the image are the one galaxy of interest and the image background noise. If an extraneous object lies within the outer, fainter, isophotes of a galaxy the software is unaware that the contaminated pixels contain contributions from an additional source and the profile generated is incorrect. For the purposes of ESP the contaminated pixels may be removed from the analysis by setting their values to bad.

6.1 Removing consistent image faults

It is well known that few, if any, CCD type detectors are perfect. Normally, it will be found that several pixels are either ‘hot’ or ‘cold’ (contain very high or very low values) no matter what is being imaged. In the case of cheaper, or older, CCDs it may be found that whole columns or rows of pixels are similiarly useless. Efforts should be made to ensure the values in these pixels are disregarded during all stages of image analysis; this may be done by identifying the pixels in question (they remain the same for a given CCD and a list is usually available at the observatory for each CCD they employ) and inserting that information into an ARD file which can be examined by the ESP application MASK. This looks at the ARD file supplied to find out which pixels are suspect and then sets them to bad in the output image thus ensuring they play no role in future analysis.

The application can be used with the following syntax:

  % mask in=p2 out=masked ardfil=^areas.dat

This looks at the ARD text file areas.dat (note the use of the ‘ ̂’ character) and then sets the pixels defined in p2 to be bad. The image so generated is output as the NDF masked. Other examples may be found in Appendix 0.

6.2 Removing other image contamination

Often, it will be found that an image of a galaxy will contain, within its isophotes, brighter areas due to satellite galaxies or other foreground and background objects. Since these do not arise at predictable sites (due to faults in the CCD or optics), it is not possible to remove these objects without a visual inspection. The simplest way to remove these is to display the image using KAPPA’s DISPLAY and to then set the offending areas of the frame to the bad value by interactive use of the KAPPA’s ARDGEN and ARDMASK applications. Unfortunately, it is usually necessary to carry out this process with all the images that are to be profiled.