Abstract : Induction is the term used to describe the fact that visual perception depends not only on the absolute lumi-nance of a patch in a visual scene, but also on the luminance of the surrounding patches. Induction can be achromatic or chromatic, depending on the presence of only shades of gray, or color, respectively. Achromatic induction can be measured through quite simple psychophysical matching experiments, whose data are well fitted by the so-called edge integration models. In this paper we will examine Rudd-Zemach's edge integration model and we will show that it can be interpreted in terms of a variational framework that shares many features with a phenomenologically-based visual perception theory recently proposed. The relevance of this result is twofold: firstly, it allows interpreting induction in terms of local contrast enhancement, secondly, it provides a validation of the variational framework in terms of quantitative measures, opening the possibility to determine important parameters of this framework.