In computer science and information theory, set redundancy compression are methods of data compression that exploits redundancy between individual data groups of a set, usually a set of similar images. It is wide used on medical and satellital images. The main methods are min-max differential, min-max predictive and centroid method.