Low-Light Image Enhancement With Semi-Decoupled Decomposition

By Shijie Hao, Xuyang Han, Yanrong Guo, Xin Xu, Meng Wang. IEEE transactions on multimedia, 2020

339 citations recorded by the source corpus.

Low-light image enhancement is important for high-quality image display and other visual applications. However, it is a challenging task as the enhancement is expected to improve the visibility of an image while keeping its visual naturalness. Retinex-based methods have well been recognized as a representative technique for this task, but they still have the following limitations. First, due to less-effective image decomposition or strong imaging noise, various artifacts can still be brought into enhanced results. Second, although the priori information can be explored to partially solve the first issue, it requires to carefully model the priori by a regularization term and usually makes the optimization process complicated. In this paper, we address these issues by proposing a novel Retinex-based low-light image enhancement method, in which the Retinex image decomposition is achieved in an efficient semi-decoupled way. Specifically, the illumination layer <inline-formula><tex-math notation="LaTeX">$I$</tex-math></inline-formula> is gradually estimated only with the input image <inline-formula><tex-math notation="LaTeX">$S$</tex-math></inline-formula> based on the proposed Gaussian Total Variation model, while the reflectance layer <inline-formula><tex-math notation="LaTeX">$R$</tex-math></inline-formula> is jointly estimated by <inline-formula><tex-math notation="LaTeX">$S$</tex-math></inline-formula> and the intermediate <inline-formula><tex-math notation="LaTeX">$I$</tex-math></inline-formula>. In addition, the imaging noise can be simultaneously suppressed during the estimation of <inline-formula><tex-math notation="LaTeX">$R$</tex-math></inline-formula>. Experimental results on several public datasets demonstrate that our method produces images with both higher visibility and better visual quality, which outperforms the state-of-the-art low-light enhancement methods in terms of several objective and subjective evaluation metrics.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/2ccae9e21e68c4cb1bf3732aadc9ad2fe2408872

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Wenhan Yang
2. Yue Huang
3. Xinghao Ding
4. Risheng Liu
5. Zhongxuan Luo
6. Xin Fan
7. Long Ma
8. Jiaao Zhang

In Media Technology

1. Wenhan Yang
2. Risheng Liu
3. Xinghao Ding
4. Yue Huang
5. Kai-Kuang Ma
6. Xin Fan
7. Zhongxuan Luo
8. Jianmin Jiang

Citations in the corpus

• Toward Fast, Flexible, and Robust Low-Light Image Enhancement (2022, positive): “The work in [10] proposed a semi-decoupled decomposition model for simultaneously improving the brightness and suppressing noises.”
• Learning a Simple Low-Light Image Enhancer from Paired Low-Light Instances (2023, methodology): “PairLIE is compared with 13 state-of-the-art LIE methods, which can be divided into the following three categories: traditional methods (SDD [9], STAR [34]), supervised approaches (MBLLEN [20], RetinexNet [32], GALDNet [30], KinD [38], DRBN [36], URetinexNet [33]), and unsupervised methods (Zero-DCE”
• Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement (2020, methodology): “sses as that in the search phase. 4.2. Comparison with State-of-the-arts We compared it qualitatively and quantitatively with twelve recently-proposed state-of-the-art methods including LIME [9], SDD [10], MBLLEN [20], GLADNet [33], RetinexNet [2], EnGAN [12], SSIENet [45], KinD [46], DeepUPE [31], ”
• URetinex-Net: Retinex-based Deep Unfolding Network for Low-light Image Enhancement (2022, methodology): “Afterward, several model-based methods whose energy functions are under Maximum a posteriori (MAP) framework are proposed [12, 13, 22, 30].”