Journal of Atmospheric and Environmental Optics, Volume. 19, Issue 3, 381(2024)
Lightweight underwater image enhancement network based on cross-scale deep distillation feature perception
Figures & Tables(7)
Fig. 1. Lightweight underwater image enhancement network structure diagram based on CSDDFP module
Fig. 2. Comparisons of parameter size and objective performance on LSUI400 dataset
Fig. 3. Visual comparisons of the mainstream methods on four benchmark datasets. (a) Input image; (b) UGAN; (c) WaterNet; (d) FUnIE_GAN; (e) UT-UIE; (f) ours; (g) ground truth
Fig. 4. Visual comparisons of the mainstream methods under insufficient illumination in deep water. (a) Input image; (b) UGAN; (c) WaterNet; (d) FUnIE_GAN; (e) UT-UIE; (f) ours; (g) ground truth
Fig. 5. Visual comparisons of the mainstream methods on texture details. (a) Input image; (b) UGAN; (c) WaterNet; (d) FUnIE_GAN; (e) UT-UIE; (f) ours; (g) ground truth
Table 1. Objective comparisons of different underwater image enhancement methods
View tableView in ArticleTable 1. Objective comparisons of different underwater image enhancement methods
Methods LSUI400 EUVP_Test515 UIEB100 OceanEx PSNR SSIM PSNR SSIM PSNR SSIM PSNR SSIM UWCNN 17.366 0.725 17.725 0.704 14.155 0.686 15.960 0.724 Cycle-GAN 18.320 0.749 17.963 0.709 17.714 0.758 21.007 0.828 SGUIE-Net 19.910 0.819 19.187 0.760 21.178 0.872 18.677 0.834 FUnIE-GAN 23.272 0.818 24.077 0.794 19.614 0.813 20.448 0.855 UGAN 25.117 0.846 23.636 0.805 21.368 0.825 22.436 0.822 UT-UIE 24.349 0.829 25.212 0.813 20.916 0.764 21.269 0.822 Ours 26.672 0.872 25.394 0.833 22.790 0.877 22.003 0.876
Table 2. Ablation studies of the cross-scale deep distillation feature perception module
View tableView in ArticleTable 2. Ablation studies of the cross-scale deep distillation feature perception module
K P/M CSDDFP SGFM[ 15 ]PPM[ 23 ]PSNR SSIM PSNR SSIM PSNR SSIM 0 1.34 25.689 0.864 25.689 0.864 25.689 0.864 1 2.99 26.439 0.870 26.012 0.866 25.902 0.865 2 4.64 26.672 0.872 26.170 0.867 26.084 0.866
Tools
Get Citation
Copy Citation Text
Xiaohua WU, Zenglu LI, Zhanghua XU, Jingchun ZHOU. Lightweight underwater image enhancement network based on cross-scale deep distillation feature perception[J]. Journal of Atmospheric and Environmental Optics, 2024, 19(3): 381
Paper Information
Category:
Received: Oct. 31, 2023
Accepted: --
Published Online: Jul. 17, 2024
The Author Email: Jingchun ZHOU (zhoujingchun03@qq.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20