In this paper, a gray-level grouping (GLG) approach exploiting the just-noticeable-difference (JND) model of the human visual system, denoted by JND-GLG, is proposed for image contrast enhancement. First, the JND model is employed to group similar gray va