most of the work was done by Osher & Fatemi , it's called "total variation denoising" without that algorithm they would never have been able to reconstruct the image