Introduction

The principal goal of restoration methods (French: méthodes de restauration) is to improve an image by using methods based on objective considerations. On the contrary, enhancement methods that we have seen (e.g. in Histogram transformations) are mostly heuristic procedures designed to improve an image for a psychophysical purpose, that is, “to look visually fine”.

Restoration methods need knowledge of the phenomenon that has degraded the image. This knowledge gives a mathematical model of the degradation. The degradation phenomenon is “inversed” to recover the original image: hence restoration methods are said to be an inverse problem (French: problème inverse). This approach usually involves formulating a criterion of goodness (French: critère de qualité) to help the search for an optimal estimation: a criterion is a mathematical function whose minimum corresponds to the best possible restoration.