EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Clean Retinex Decomposition Network for Low-Light Image Enhancement

Yukun Ma, Minghui Qiao, Xiaowen Li, Xinze Li, He Wang

Neural Processing Letters · 20.07.2026 · 58 (5)

DOI: 10.1007/s11063-026-11875-w ↗

Kurz gefasst

Fachartikel in der Zeitschrift Neural Processing Letters, online erschienen am 20.07.2026. Verfasst von Yukun Ma und 4 weiteren Autor:innen. Die Arbeit nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: the Training Program for Young Backbone Teachers of Higher Education Institutions in Henan Province China, the Henan Provincial Scientifc and Technological Research Project, the Henan Provincial Natural Science Foundation Project.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Images captured in low-light environments often suffer from various complex degradations. The main task of low-light image enhancement (LLIE) is to improve the brightness and restore the details of the image, thereby producing images that are more in line with human perception. Retinex-based methods decompose images into illumination and reflectance components, and enhance images by adjusting illumination and restoring reflectance details. However, the presence of noise in images usually leads to inaccurate decomposition, producing results that deviate from realistic appearances. In order to prevent the influence of noise on the subsequent decomposition tasks, we first denoise the low-light images, perform Retinex decomposition on the denoised images, and try to establish paired low-light images to maintain the consistency of the reflectance components. Specifically, we propose a Clean Decomposition Network (CDNet), an unsupervised learning method that first performs denoising on low-light image pairs to prevent loss of details and color deviation, and then performs Retinex decomposition, and guides network optimization through a carefully designed denoising network and loss function. Extensive experiments on eight public datasets, including LOL and SICE, demonstrate that our method achieves superior enhancement performance, is highly competitive with state-of-the-art unsupervised methods, and achieves comparable levels to supervised methods on certain metrics and datasets.

Förderung

  • the Training Program for Young Backbone Teachers of Higher Education Institutions in Henan Province China
  • the Henan Provincial Scientifc and Technological Research Project
  • the Henan Provincial Natural Science Foundation Project

Zitieren

APA

Ma, Y., Qiao, M., Li, X., Li, X., & Wang, H. (2026). Clean Retinex Decomposition Network for Low-Light Image Enhancement. Neural Processing Letters, 58(5). https://doi.org/10.1007/s11063-026-11875-w

Vancouver

Ma Y, Qiao M, Li X, Li X, Wang H. Clean Retinex Decomposition Network for Low-Light Image Enhancement. Neural Processing Letters. 2026;58(5). doi:10.1007/s11063-026-11875-w

BibTeX
@article{Ma2026,
  author = {Ma, Yukun and Qiao, Minghui and Li, Xiaowen and Li, Xinze and Wang, He},
  title = {Clean Retinex Decomposition Network for Low-Light Image Enhancement},
  journal = {Neural Processing Letters},
  year = {2026},
  volume = {58},
  number = {5},
  doi = {10.1007/s11063-026-11875-w}
}
Kostenlos starten →