Bayesian image restoration in astronomy: Application to images of the recent collision of comet Shoemaker-Levy 9 with Jupiter

In this work, we examine simple to complex methods proposed within the Bayesian paradigm to perform image restoration in astronomy. We start by describing the classical conditional and simultaneous autoregressions, then we move on to study how to incorporate smoothness constraints to the classical Richardson-Lucy restoration method and also how to modify the image scale and define prior models on other scales than the linear one. Finally, we compare those models on images of Jupiter after the impacts of the fragments of the comet Shoemaker-Levy 9 at two wavelengths. (C) 1995 John Wiley & Sons, Inc.