Enhancing Low-Light Color Image via L ₀ Regularization and Reweighted Group Sparsity

Abstract

Classic Retinex model based low-light image enhancement methods ignored the interference of noise, which causes annoying artifacts. In this paper, we propose to estimate the illumination, reflectance and suppress the noise in a whole framework. Instead of using the L 1 norm to constrain the piece-wise smoothness, we utilize the L 0 norm to preserve the structure of the illumination map and remove the intensive noise. The clean reflectance is obtained via a novel group sparsity regularization to preserve the small scale details. Instead of using a zero-mean model for all sparse coefficients, we propose to adaptively estimate the mean of each coefficient according to the statistical characteristics of the image content. A re-weighting scheme is introduced to adjust how close the estimated patch is to the mean value. In addition, based on the observation that the noise levels in different color channels are different, the noise variance in each channel is estimated and updated during the model optimization process. Experimental results show that the proposed method outperforms the compared schemes in terms of both objective quality and visual quality.INDEX TERMS Low-light image enhancement, L 0 sparsity, reweighted group sparsity, noise model.

Unsupervised Image Enhancement Method Based on Attention Map Network Guidance and Attention Mechanism

“…Zhang et al [13] proposed a variational Retinex model based on global sparse gradient (GSG) guidance, which incorporates GSG Regularization to preserve the structural information and edge details of images, thereby improving the enhancement effect based on the Retinex principle. Song et al [14] introduced an image enhancement method based on L0 regularization and re-weighted group sparsity (RGS), where L0 Regularization can better preserve image details, while RGS can better retain image colors.…”