Dual-Stage Low-Light Image Enhancement Network Based on HVI Color Space Transformer

doi:10.19678/j.issn.1000-3428.0253485

Abstract

Abstract:

Low-Light Image Enhancement (LLIE) is crucial in computer vision by restoring rich visual information from corrupted low-light images. However, existing LLIE methods often suffer from color bias due to color space and typically fail to balance denoising and color fidelity within a single-stage framework. To address these challenges, this study introduces a novel Dual-stage HVI-based Transformer Network (DHTNet) for LLIE. DHTNet significantly improves the quality of low-light images by decoupling I and HV maps, enabling their independent yet synergistic optimization within the HVI color space. In the first stage, a hierarchical Transformer network equipped with an Adaptive Guidance Interaction Module (AGIM) models long-range dependencies between I and HV features to achieve global noise suppression and accurate color calibration. In the second stage, the Multi-Scale Enhanced Synergistic Attention (MESA) module enhances localized color and detail representation through synergistic optimization across the I and HV branches. This dual-stage framework effectively addresses the limitations of existing LLIE methods by retaining complex image details while enhancing visual realism. Experimental results show that DHTNet achieves the highest Peak Signal-to-Noise Ratio (PSNR) on both the SICE and SID datasets, outperforming the second-best model by 0.717 dB and 1.897 dB, respectively. Furthermore, DHTNet attains PSNR values of 28.756 dB, 24.683 dB, and 25.950 dB on the LOLv1, LOLv2-Real, and LOLv2-Synthetic datasets, respectively, consistently outperforming existing models such as Retinexformer and CIDNet.

Key words: Low-Light Image Enhancement (LLIE), HVI color space, dual-stage, adaptive guidance interaction, attention mechanism

摘要：

低光图像增强(LLIE)是计算机视觉领域的一项关键技术, 旨在从质量退化的低光图像中恢复丰富的视觉信息。然而, 现有方法因对颜色空间敏感, 常出现色彩偏差问题, 且在单阶段框架下难以兼顾噪声抑制与色彩保真度。针对上述问题, 提出一种基于HVI颜色空间变换的双阶段网络(DHTNet)。该网络模型通过解耦I和HV图, 实现对两者的独立优化与协同增强, 从而显著提升低光图像的视觉质量。第一阶段采用集成了自适应引导交互模块(AGIM)的分层Transformer网络建模了I和HV特征之间的长距离依赖关系, 以实现全局噪声抑制和精准色彩校正。第二阶段引入多尺度增强协同注意力(MESA)模块, 通过跨分支的协同优化机制增强局部色彩与细节特征。该双阶段框架在保留复杂图像结构的同时有效提升了视觉真实性, 解决了现有LLIE方法的局限性。在SICE和SID数据集上的实验结果表明: DHTNet的峰值信噪比(PSNR)均达到最高水平, 较次优模型提升0.717 dB和1.897 dB; 在LOLv1、LOLv2-Real和LOLv2-Synthetic数据集上, PSNR分别达到28.756 dB、24.683 dB和25.950 dB, 性能优于Retinexformer、CIDNet等对比模型。

关键词: 低光图像增强, HVI颜色空间, 双阶段, 自适应引导交互, 注意力机制

LI Yamin, XIANG Wen, CHAI Li, XIANG Yao. Dual-Stage Low-Light Image Enhancement Network Based on HVI Color Space Transformer[J]. Computer Engineering, 2026, 52(9): 228-241.

李亚敏, 向稳, 柴丽, 向尧. 基于HVI颜色空间变换的双阶段低光图像增强网络[J]. 计算机工程, 2026, 52(9): 228-241.

/ Recommend / Download Citations

URL: https://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0253485

https://www.ecice06.com/EN/Y2026/V52/I9/228

Figures/Tables 17

Fig.1 Overall framework of DHTNet

Fig.2 AGIM structure

Fig.3 MESA structure

Fig.4 Visual comparison of LLIE results on LOLv1 dataset

Fig.5 Visual comparison of LLIE results on LOLv2-Real dataset

Fig.6 Visual comparison of LLIE results on LOLv2-Synthetic dataset

Fig.7 Visual comparison of LLIE results on five unpaired dataset

Fig.8 Visual effect on dual-stage processing

Fig.9 Visual comparison of object detection results on the ExDark dataset using the proposed method under low-light conditions and low-light enhancement conditions

Fig.10 Qualitative visual results under two extreme scene diagrams

References 46

1	ALAGARSAMY M , KASINATHAN P , MANICKAM G , et al. IoT based E-vehicle monitoring system using sensors and imaging processing algorithm. International Journal of Reconfigurable and Embedded Systems (IJRES), 2022, 11 (2): 196. doi: 10.11591/ijres.v11.i2.pp196-204
2	WANG J L, ZHUANG W J, SHANG D. Light enhancement algorithm optimization for autonomous driving vision in night scenes based on YOLACT++[C]//Proceedings of the 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS). Washington D.C., USA: IEEE Press, 2022: 417-423.
3	GUL F , RAHIMAN W , ALHADY S S N . A comprehensive study for robot navigation techniques. Cogent Engineering, 2019, 6 (1): 1632046. doi: 10.1080/23311916.2019.1632046
4	HASHMI K A, KALLEMPUDI G, STRICKER D, et al. FeatEnHancer: enhancing hierarchical features for object detection and beyond under low-light vision[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Washington D.C., USA: IEEE Press, 2023: 6702-6712.
5	PIZER S M , AMBURN E P , AUSTIN J D , et al. Adaptive histogram equalization and its variations. Computer Vision, Graphics, and Image Processing, 1987, 39 (3): 355- 368. doi: 10.1016/S0734-189X(87)80186-X
6	WEI C, WANG W J, YANG W H, et al. Deep retinex decomposition for low-light enhancement[EB/OL]. [2025-11-07]. https://arxiv.org/abs/1808.04560.
7	范政巍, 苌道方, 满星妤, 等. 双阶段Retinex焊缝低光图像增强与缺陷检测[J/OL]. 计算机工程: 1-13[2025-11-07]. https://doi.org/10.19678/j.issn.1000-3428.0252732.
	FAN Z W, CHANG D F, MAN X Y, et al. Two-stage Retinex weld seam low-light image enhancement and defect detection[J/OL]. Computer Engineering: 1-13[2025-11-07]. https://doi.org/10.19678/j.issn.1000-3428.0252732. (in Chinese)
8	GUO C L, LI C Y, GUO J C, et al. Zero-reference deep curve estimation for low-light image enhancement[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2020: 1777-1786.
9	WANG T, ZHANG K H, SHEN T R, et al. Ultra-high-definition low-light image enhancement: a benchmark and Transformer-based method[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Palo Alto, USA: AAAI Press, 2023: 2654-2662.
10	WANG Z D, CUN X D, BAO J M, et al. Uformer: a general U-shaped Transformer for image restoration[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2022: 17662-17672.
11	ZAMIR S W, ARORA A, KHAN S, et al. Restormer: efficient Transformer for high-resolution image restoration[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2022: 5718-5729.
12	CAI Y H, BIAN H, LIN J, et al. Retinexformer: one-stage retinex-based Transformer for low-light image enhancement[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Washington D.C., USA: IEEE Press, 2023: 12470-12479.
13	YAN Q S, FENG Y X, ZHANG C, et al. HVI: a new color space for low-light image enhancement[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2025: 5678-5687.
14	MUSTAFA W A , KADER M M M A . A review of histogram equalization techniques in image enhancement application. Journal of Physics: Conference Series, 2018, 1019 (1): 012026.
15	AL-AMEEN Z . Contrast enhancement for color images using an adjustable contrast stretching technique. International Journal of Computing, 2018, 74- 80. doi: 10.47839/ijc.17.2.993
16	SHARIF S M A, MYRZABEKOV A, KHUJAEV N, et al. Learning optimized low-light image enhancement for edge vision tasks[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Washington D.C., USA: IEEE Press, 2024: 6373-6383.
17	ZHANG Y H, ZHANG J W, GUO X J. Kindling the darkness: a practical low-light image enhancer[C]// Proceedings of the 27th ACM International Conference on Multimedia. New York, USA: ACM Press, 2019: 1632-1640.
18	WANG Y F, WAN R J, YANG W H, et al. Low-light image enhancement with normalizing flow[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Palo Alto, USA: AAAI Press, 2022: 2604-2612.
19	JIANG Y F , GONG X Y , LIU D , et al. EnlightenGAN: deep light enhancement without paired supervision. IEEE Transactions on Image Processing, 2021, 30, 2340- 2349. doi: 10.1109/TIP.2021.3051462
20	XU X G, WANG R X, FU C W, et al. SNR-aware low-light image enhancement[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2022: 17693-17703.
21	FU Z Q, YANG Y, TU X T, et al. Learning a simple low-light image enhancer from paired low-light instances[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2023: 22252-22261.
22	李紫薇, 刘金龙, 杨慧珍, 等. 基于深度学习的低照度图像增强算法综述. 应用光学, 2024, 45 (6): 1095- 1107.
	LI Z W , LIU J L , YANG H Z , et al. Review of low-illuminance image enhancement algorithm based on deep learning. Journal of Applied Optics, 2024, 45 (6): 1095- 1107.
23	王非凡, 陈希爱, 任卫红, 等. 基于图像自适应增强的低照度目标检测算法. 计算机工程, 2024, 50 (10): 352- 361. doi: 10.19678/j.issn.1000-3428.0068407
	WANG F F , CHEN X A , REN W H , et al. Low-illumination object-detection algorithm based on image adaptive enhancement. Computer Engineering, 2024, 50 (10): 352- 361. doi: 10.19678/j.issn.1000-3428.0068407
24	宋瑞霞, 李达, 王小春. 基于HSI色彩空间的低照度图像增强算法. 图学学报, 2017, 38 (2): 217- 223.
	SONG R X , LI D , WANG X C . Low illumination image enhancement algorithm based on HSI color space. Journal of Graphics, 2017, 38 (2): 217- 223.
25	WANG T , ZHANG K H , ZHANG Y , et al. LLDiffusion: Learning degradation representations in diffusion models for low-light image enhancement. Pattern Recognition, 2025, 166, 111628. doi: 10.1016/j.patcog.2025.111628
26	XU L T , HU C H , ZHANG B , et al. Swin Transformer and ResNet based deep networks for low-light image enhancement. Multimedia Tools and Applications, 2024, 83 (9): 26621- 26642.
27	LAND E H . The retinex theory of color vision. Scientific American, 1977, 237 (6): 108- 128. doi: 10.1038/scientificamerican1277-108
28	KINGMA D P, BA J. Adam: a method for stochastic optimization[EB/OL]. [2025-11-07]. https://arxiv.org/abs/1412.6980.
29	LOSHCHILOV I, HUTTER F. SGDR: stochastic gradient descent with warm restarts[EB/OL]. [2025-11-07]. https://arxiv.org/abs/1608.03983.
30	WANG Z , BOVIK A C , SHEIKH H R , et al. Image quality assessment: from error visibility to structural similarity. IEEE Transactions on Image Processing, 2004, 13 (4): 600- 612. doi: 10.1109/TIP.2003.819861
31	ZHANG R, ISOLA P, EFROS A A, et al. The unreasonable effectiveness of deep features as a perceptual metric[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2018: 586-595.
32	MITTAL A , MOORTHY A K , BOVIK A C . No-reference image quality assessment in the spatial domain. IEEE Transactions on Image Processing, 2012, 21 (12): 4695- 4708. doi: 10.1109/TIP.2012.2214050
33	MITTAL A , SOUNDARARAJAN R , BOVIK A C . Making a "completely blind" image quality analyzer. IEEE Signal Processing Letters, 2012, 20 (3): 209- 212.
34	YANG W H , WANG W J , HUANG H F , et al. Sparse gradient regularized deep retinex network for robust low-light image enhancement. IEEE Transactions on Image Processing, 2021, 30, 2072- 2086. doi: 10.1109/TIP.2021.3050850
35	CAI J R , GU S H , ZHANG L . Learning a deep single image contrast enhancer from multi-exposure images. IEEE Transactions on Image Processing, 2018, 27 (4): 2049- 2062. doi: 10.1109/TIP.2018.2794218
36	CHEN C, CHEN Q F, XU J, et al. Learning to see in the dark[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2018: 3291-3300.
37	GUO X J , LI Y , LING H B . LIME: low-light image enhancement via illumination map estimation. IEEE Transactions on Image Processing, 2016, 26 (2): 982- 993.
38	WANG S H , ZHENG J , HU H M , et al. Naturalness preserved enhancement algorithm for non-uniform illumination images. IEEE Transactions on Image Processing, 2013, 22 (9): 3538- 3548. doi: 10.1109/TIP.2013.2261309
39	LEE C , LEE C , KIM C S . Contrast enhancement based on layered difference representation of 2D histograms. IEEE Transactions on Image Processing, 2013, 22 (12): 5372- 5384. doi: 10.1109/TIP.2013.2284059
40	VONIKAKIS V , KOUSKOURIDAS R , GASTERATOS A . On the evaluation of illumination compensation algorithms. Multimedia Tools and Applications, 2018, 77 (8): 9211- 9231. doi: 10.1007/s11042-017-4783-x
41	ZHENG S, MA Y L, PAN J Q, et al. Low-light image and video enhancement: a comprehensive survey and beyond[EB/OL]. [2025-11-07]. https://arxiv.org/abs/2212.10772.
42	LIU R S, MA L, ZHANG J A, et al. Retinex-inspired unrolling with cooperative prior architecture search for low-light image enhancement[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2021: 10556-10565.
43	GUO X J , HU Q M . Low-light image enhancement via breaking down the darkness. International Journal of Computer Vision, 2023, 131 (1): 48- 66. doi: 10.1007/s11263-022-01667-9
44	WU W H, WENG J, ZHANG P P, et al. URetinex-Net: retinex-based deep unfolding network for low-light image enhancement[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2022: 5891-5900.
45	LOH Y P , CHAN C S . Getting to know low-light images with the exclusively dark dataset. Computer Vision and Image Understanding, 2019, 178, 30- 42. doi: 10.1016/j.cviu.2018.10.010
46	LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft COCO: common objects in context[C]//Proceedings of European Conference on Computer Vision. Berlin, Germany: Springer, 2014: 740-755.

[1]	ZHAI Sheping, MA Mengyao, ZHANG Wenjing, YANG Rui. Knowledge Graph Completion Integrating Relational Context Semantics [J]. Computer Engineering, 2026, 52(9): 113-122.
[2]	HAN Yanling, ZHU Xiaojun, WANG Jing, PAN Haiyan, ZHANG Yun. Passive Microwave Sea Ice Thickness Inversion Based on SIT-TransNET [J]. Computer Engineering, 2026, 52(8): 390-399.
[3]	ZHAO Tao, DONG Lihong, QIN Yi. Underground Personnel Detection Algorithm Combining Attention and Lightweight Network [J]. Computer Engineering, 2026, 52(8): 400-411.
[4]	JIN Jing, HU Chudi, CHEN Gang. Medical Image Segmentation Method Based on Multi-Scale Self-Attention Transformer [J]. Computer Engineering, 2026, 52(7): 199-209.
[5]	LI Zhongwei, LI Keyi, LIU Xin. Sequential Recommendation Algorithm Based on Data Augmentation and Integrated Self-Attention [J]. Computer Engineering, 2026, 52(7): 94-106.
[6]	LI Jinru, PAN Qingxian, GAO Zhaolong, WEI Kai, FAN Zequn. Crowdsourcing Task Recommendation Method Based on Collaborative Knowledge Graph and Hybrid Neural Network [J]. Computer Engineering, 2026, 52(7): 155-167.
[7]	LIANG Ziyi, WANG Zihao, LIU Tianquan, LI Liping, ZHU Yuanfei, LU Cunyue. PASRGAN: A Super-Resolution Pathological Image Generation Algorithm Based on Parallel Attention Mechanism [J]. Computer Engineering, 2026, 52(7): 189-198.
[8]	HE Mingyan, LI Siyuan, LIU Peng, HUANG Jianhua. Opponent Modeling Based on Cross-Attention Diffusion Model [J]. Computer Engineering, 2026, 52(7): 76-85.
[9]	LIU Xiangbin, FANG Cheng, LIU Shuai. Real-Time Semantic Segmentation Algorithm Based on Multi-Scale Feature Fusion and Polarized Self-Attention Mechanism [J]. Computer Engineering, 2026, 52(7): 210-219.
[10]	YANG Xinyi, HE Xiaohai, TENG Qizhi, CHEN Honggang, QING Linbo, WU Xiaohong. Cross-Modal Named Entity Recognition Based on Semantic Fusion [J]. Computer Engineering, 2026, 52(7): 345-353.
[11]	ZHOU Lijun, ZHANG Junran, WANG Kaiyuan, XIANG Junlian. Drug Recommendation Based on Graph-Enhanced Attention Integrating Clinical Features of Patients [J]. Computer Engineering, 2026, 52(6): 314-325.
[12]	DAI Yinqiao, XIAO Wulong, LI Bailin, LI Li. Lettuce Core Detection Algorithm Based on Improved YOLOv5s [J]. Computer Engineering, 2026, 52(6): 352-364.
[13]	ZENG An, ZHENG Jiayu, PAN Dan, ZHAO Jingliang, HUANG Xingqing. Centerline Tracking Algorithm for Aortic Dissection Based on Deep Reinforcement Learning [J]. Computer Engineering, 2026, 52(6): 414-424.
[14]	YU Mengyuan, LIU Xiangyang. Ship Detection Method Based on Multimodal Visible and Infrared Image Fusion [J]. Computer Engineering, 2026, 52(6): 278-287.
[15]	WANG Yongqi, WANG Lei. Multimodal Sentiment Recognition Based on Cross-Modal Enhancement and Time-Step Gating [J]. Computer Engineering, 2026, 52(6): 258-267.

Please choose a citation manager

Content to export