by E. Thiebaut, C. Pichon, R. Cannon, J. Pichon
Since observed images or cubes are blurred by the instrument and transfer medium, image or volume reconstruction is a key problem in observational astronomy. Similarly, the creation of mock data set via discrete Monte Carlo simulations introduces some bluring and noise that needs to be taken care of. Understanding the fundamental problems underlying the deconvolution (noise amplification) and the way to solve for them (regularization) is the purpose of MAAD: a Multiresolution Algorithm for Automatic Deblurring.
It may be used for
recovering information from pseudo images or
as a tool to find the optimal level of smoothing applied to a given discrete realisation.