Adaptive Variational Model for Contrast Enhancement of Low-Light Images

Contrast enhancement plays an important role in image/video processing and computer vision applications. Its main purpose is to adjust the image intensity to enhance the quality and features of the image. In this paper, we propose a simple and efficient adaptive variational model for contrast enhancement for partially shaded low-light images. The key idea of this adaptive approach is to employ the maximum image of the RGB color channels as a classifier to divide the image domain into the relatively bright and dim parts, and then use different fitting terms for each part such that the bright pixels are preserved as close as possible to the original ones while the dim pixels are boosted with brightness and contrast-level parameters to adjust the degree of the strength. With this adaptivity, one can find that the proposed model considerably improves upon the existing variational models in the literature. In this paper, the existence and uniqueness of the minimizer for the variational minimization problem is established. The split Bregman method is used to accomplish an efficient numerical implementation of the adaptive variational model. Moreover, a number of numerical experiments and comparisons with other popular enhancement methods are conducted to demonstrate the high performance of the newly proposed method.