Back to Volume
Paper: Evaluation of Image Restoration algorithms applied to HST images
Volume: 61, Astronomical Data Analysis Software and Systems III
Page: 304
Authors: Busko, I.
Abstract: Very few research was conducted up to now to quantify the photometric behavior of image restoration algorithms when applied to HST imagery of stellar fields. Of note are the works by Cohen (1991, Astron.J. , 101 ,734) and Thomson et al. (1992, Mon.Not.R.astr.Soc. , 259 ,104). Both of them, however, focus on the MEMSYS implementation of the Maximum Entropy technique, and almost do not touch other methods. We conducted a systematic investigation of algorithms available in the Space Telescope Science Data Analysis System: Lucy, Wiener filter and sigma-CLEAN, as well as variants of the VanCittert iteration and the Maximum Entropy technique. Experiments involved the use of simulated images as observed by HST. The analysis took into account criteria as photometric linearity and precision, as well as a global improvement factor which is independent of image content. Results point to the existence of a trade-off between a particular algorithm's linearity, and its ability to control high-frequency noise buildup, as well as its capability to decrease confusion between neighbor star images. Tests involving several independent realizations of the same image showed that restoration does not improve, but does not degrade either, photometric precision when compared with precision achievable in unrestored images.
Back to Volume