Image enhancement in turbid water using multiscale weighted features and attention mechanisms

Abstract

In this study, we propose a method to improve the image quality in turbid water by combining multiscale weighted features and attention mechanisms, effectively addressing issues such as reduced contrast and color casts. Initially, global color correction was employed using the grayscale space algorithm. Then, a preprocessed dataset was input into an improved Shallow-UWnet network for feature learning. The model included a multiscale weighted feature fusion module to extract global and local features and a parallel attention module to enhance key details and suppress noise in images. To simulate natural turbid water, a dataset of underwater images with different turbidities was constructed. Our method effectively corrected the color of images in different turbid water conditions and significantly enhanced image quality. In the evaluation on the custom dataset, performance improved by 13%–25% across standard image quality metrics compared with the second-best method. Additionally, we conducted generalization performance tests on the EUVP Dark, UIEBD, and UFO-120 datasets, validating the excellent enhancement effect of our method across different types of images. These results highlight the promising potential of our method for applications in underwater detection, rescue, and marine research.

Keywords

Turbid water / Underwater image enhancement / Improved Shallow-UWnet network / Multiscale weighted features / Attention mechanism

References

[1] Bakht AB, Jia ZK, Din MU, Akram W, Saoud LS, Seneviratne L, et al.. MuLA-GAN: multi-level attention GAN for enhancedunderwatervisibility. Ecol Informatics, 2024, 81 102631

[2] Drews P, do Nascimento E, Moraes F, Botelho S, Campos M (2013) Transmission estimation in underwater single images. In: Proceedings of the IEEE International Conference on Computer Vision Workshops. IEEE, pp 825–830. https://doi.org/10.1109/ICCVW.2013.113

[3] Duarte A, Codevilla F, Gaya JD, Botelho SSC (2016) A dataset to evaluate underwater image restoration methods. In: OCEANS 2016–Shanghai. IEEE, pp 1–6. https://doi.org/10.1109/OCEANSAP.2016.7485524

[4] Hu K, Weng CH, Zhang YW, Jin JL, Xia QF. An overview of underwater vision enhancement: fromtraditional methods to recent deep learning. J Mar Sci Eng, 2022, 10(2 241

[5] Islam MJ, Luo PG, Sattar J (2020a) Simultaneous enhancement and super-resolution of underwater imagery for improved visual perception. Preprint at arXiv:2002.01155

[6] Islam MJ, Xia YY, Sattar J. Fast underwater image enhancement for improved visualperception. IEEE Robot Autom Lett, 2020, 5(2): 3227-3234

[7] Jiang JX, Ye T, Bai JB, Chen SX, Chai WH, Jun S et al (2023) Five A⁺ Network: you only need 9k parameters for underwater image enhancement. Preprint at arXiv:2305.08824

[8] Johnson J, Alahi A, Li FF (2016) Perceptual losses for real-time style transfer and super-resolution. In: Leibe B et al (eds) Computer Vision–ECCV 2016. Lecture notes in computer science, vol 9906. Springer, Cham, pp 694–711. https://doi.org/10.1007/978-3-319-46475-6_43

[9] Li CY, Guo CL, Ren WQ, Cong RM, Hou JH, Kwong S, et al.. An underwater image enhancement benchmark dataset andbeyond. IEEE Trans Image Process, 2019, 29: 4376-4389

[10] Li HY, Wang H, Zhang Y, Li L, Ren P. Underwater image captioning: challenges, models, anddatasets. ISPRS J Photogramm Remote Sens, 2025, 220: 440-453

[11] Li KQ, Wu L, Qi Q, Liu WJ, Gao X, Zhou LQ, et al.. Beyond single reference for training: underwater image enhancement via comparative learning. IEEE Trans Circuits Syst Video Technol, 2022, 33(6): 2561-2576

[12] Lin S, Ning ZM, Zhang RH. Modified optical model and optimized contrast for underwater image restoration. Opt Commun, 2025, 574 130942

[13] Liu HM, Ding Y, Zeng H, Pu HY, Luo J, Fan B. A cascaded multimodule image enhancement framework forunderwater visual perception. IEEE Trans Neural Netw Learn Syst, 2025, 36(4): 6286-6298

[14] Liu HM, Zhang Q, Hu YF, Zeng H, Fan B. Unsupervised multi-expert learning model for underwater image enhancement. IEEE-CAA J Automatica Sin, 2024, 11(3): 708-722

[15] Liu SB, Fan HJ, Lin S, Wang Q, Ding ND, Tang YD. Adaptive learning attention network for underwater image enhancement. IEEE Robot Autom Lett, 2022, 7(2): 5326-5333

[16] Ma HP, Huang JY, Shen CX, Jiang ZH. Retinex-inspired underwater image enhancement withinformation entropy smoothing and non-uniform illumination priors. Pattern Recognit, 2025, 162 111411

[17] Naik A, Swarnakar A, Mittal KAAAI. Shallow-UWnet: compressed model for underwaterimage enhancement (student abstract). Proceedings of the AAAI Conference on ArtificialIntelligence, 2021

[18] Peng LT, Zhu CL, Bian LH. U-shape transformer for underwater image enhancement. IEEE Trans Image Process, 2023, 32: 3066-3079

[19] Shen X, Zhao XN, Luo R (2023) Semantic embedded deep neural network: a generic approach to boost multi-label image classification performance. Preprint at arXiv:2305.05228

[20] Song W, Wang Y, Huang DM, Tjondronegoro D (2018) A rapid scene depth estimation model based on underwater light attenuation prior for underwater image restoration. In: Hong R et al (eds) Advances in Multimedia Information Processing–PCM 2018. Lecture notes in computer science, vol 11164. Springer, Cham, pp 678–688. https://doi.org/10.1007/978-3-030-00776-8_62

[21] Sun BY, Mei YP, Yan N, Chen YY. Umgan: underwater image enhancement network for unpaired image-to-image translation. J Mar Sci Eng, 2023, 11(2 447

[22] Wang H, Sun SX, Chang LB, Li HY, Zhang WW, Frery AC, et al.. Inspiration: a reinforcement learning-based humanvisual perception-driven image enhancement paradigm for underwater scenes. Eng Appl Artif Intell, 2024, 133 108411

[23] Wang H, Sun SX, Ren P. Underwater color disparities: cues for enhancing underwaterimages toward natural color consistencies. IEEE Trans Circuits Syst Video Technol, 2023, 34(2): 738-753

[24] Wang H, Sun SX, Ren P. Meta underwater camera: a smart protocol for underwaterimage enhancement. ISPRS J Photogramm Remote Sens, 2023, 195: 462-481

[25] Wang QL, Wu BG, Zhu PF, Li PH, Zuo WM, Hu QH (2020) ECA-Net: efficient channel attention for deep convolutional neural networks. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, pp 11531–11539. https://doi.org/10.1109/CVPR42600.2020.01155

[26] Xiang D, Zhou ZB, Yang WL, Wang HH, Gao P, Xiao MM, et al.. A fusion framework with multi-scale convolution and triple-branch cascaded transformer for underwater image enhancement. Opt Lasers Eng, 2025, 184 108640

[27] Zhang H, Li XC, Yao GP, Gong LX, Tian YB. A pulse delay control method by external trigger to underwater illumination. J Nanoelectron Optoelectron, 2022, 17(8): 1185-1194

[28] Zhang WD, Liu QM, Lu HM, Wang JP, Liang J. Underwater image enhancement via wavelet decomposition fusion of advantage contrast. IEEE Trans Circuits Syst Video Technol, 2025, 35: 7807-7820

[29] Zhang XH, Zou YX, Shi W (2017) Dilated convolution neural network with LeakyReLU for environmental sound classification. In: 2017 22nd International Conference on Digital Signal Processing (DSP). IEEE, pp 1–5. https://doi.org/10.1109/ICDSP.2017.8096153

[30] Zhao HY, Kong XT, He JW, Qiao Y, Dong C (2020) Efficient image super-resolution using pixel attention. In: Bartoli A et al (eds) Computer Vision–ECCV 2020 Workshops. Lecture notes in computer science, vol 12537. Springer, Cham, pp 56–72. https://doi.org/10.1007/978-3-030-67070-2_3

[31] Zheng HM, Wu YW, Su YG. Underwater image enhancement based on variational image decomposition. J Opt, 2025

[32] Zhu JY, Park T, Isola P, Efros AA (2017) Unpaired image-to-image translation using cycle-consistent adversarial networks. In: 2017 IEEE International Conference on Computer Vision (ICCV). IEEE, pp 2242–2251. https://doi.org/10.1109/ICCV.2017.244

Funding

Key Research Projects of Qingdao Science and Technology Plan(22-3-3-hygg-30-hy)

Natural Science Foundation of Shandong Province(ZR2022ZD38)

RIGHTS & PERMISSIONS

The Author(s)