Visibility restoration for real-world hazy images via improved physical model and Gaussian total variation

doi:10.1007/s11704-023-3394-0

Front. Comput. Sci.

2024, Vol. 18

Issue (1) : 181708 https://doi.org/10.1007/s11704-023-3394-0

Image and Graphics

Visibility restoration for real-world hazy images via improved physical model and Gaussian total variation

Chuan LI^1,², Enping HU³, Xinyu ZHANG³, Hao ZHOU¹, Hailing XIONG⁴, Yun LIU³(

)

¹. College of Computer and Information Science, Southwest University, Chongqing 400715, China
². School of Big Data and Intelligent Engineering, Chongqing College of International Business and Economics, Chongqing 401520, China
³. College of Artificial Intelligence, Southwest University, Chongqing 400715, China
⁴. College of Electronic and Information Engineering, Southwest University, Chongqing 400715, China

Download: PDF(4051 KB) HTML
Export: BibTeX | EndNote | Reference Manager | ProCite | RefWorks

Corresponding Author(s): Yun LIU

Just Accepted Date: 04 December 2023 Issue Date: 18 January 2024

Cite this article:

Chuan LI,Enping HU,Xinyu ZHANG, et al. Visibility restoration for real-world hazy images via improved physical model and Gaussian total variation[J]. Front. Comput. Sci., 2024, 18(1): 181708.

URL:

https://academic.hep.com.cn/fcs/EN/10.1007/s11704-023-3394-0
https://academic.hep.com.cn/fcs/EN/Y2024/V18/I1/181708

Fig.1 Flowchart of the proposed approach

Fig.2 Visibility restoration results by different methods. (a) Hazy images; (b) CAP; (c) BCPDCP; (d) DehazeNet; (e) MSCNN; (f) PDN; (g) PMD-Net; (h) ours

Methods	CAP	BCPDCP	DehazeNet	MSCNN	PDN	PMD-Net	Our
NIQE $↓$	3.9024	3.8587	4.1075	3.9281	3.7790	4.6733	3.7512
FADE $↓$	1.3131	0.6876	0.8096	0.9375	0.6908	1.1059	0.7674
R $↑$	1.2924	1.6574	1.4014	1.5191	1.7824	1.3622	2.9249

Tab.1 Quantitative comparisons of 100 real-world images

Fig.3 Effects of three weight parameters (

α

β

δ

)

Fig.4 Comparisons of noise suppression. (a) Hazy image; (b) noisy hazy image; (c) CAP; (d) BCPDCP; (e) DehazeNet; (f) MSCNN; (g) PDN; (h) PMD-Net; (i) ours; (j) noise map

Fig.5 Average convergence curve of 100 real-world images

Fig.6 Extended applications on low-light and sandstorm degraded images

1	Q, Zhu J, Mai L Shao . A fast single image haze removal algorithm using color attenuation prior. IEEE Transactions on Image Processing, 2015, 24( 11): 3522–3533
2	C, Li C, Yuan H, Pan Y, Yang Z, Wang H, Zhou H Xiong . Single image dehazing based on improved bright channel prior and dark channel prior. Electronics, 2023, 12( 2): 299
3	W, Ren S, Liu H, Zhang J, Pan X, Cao M H Yang . Single image dehazing via multi-scale convolutional neural networks. In: Proceedings of the 14th European Conference on Computer Vision. 2016, 154−169
4	B, Cai X, Xu K, Jia C, Qing D Tao . DehazeNet: an end-to-end system for single image haze removal. IEEE Transactions on Image Processing, 2016, 25( 11): 5187–5198
5	D, Yang J Sun . Proximal Dehaze-Net: a prior learning-based deep network for single image dehazing. In: Proceedings of the 15th European Conference on Computer Vision. 2018, 729−746
6	Ye T, Zhang Y, Jiang M, Chen L, Liu Y, Chen S, Chen E. Perceiving and modeling density for image dehazing. In: Proceedings of the 17th European Conference on Computer Vision. 2022, 130−145
7	S, Hao X, Han Y, Guo X, Xu M Wang . Low-light image enhancement with semi-decoupled decomposition. IEEE Transactions on Multimedia, 2020, 22( 12): 3025–3038
8	L K, Choi J, You A C Bovik . Referenceless prediction of perceptual fog density and perceptual image defogging. IEEE Transactions on Image Processing, 2015, 24( 11): 3888–3901
9	N, Hautière J P, Tarel D, Aubert E Dumont . Blind contrast enhancement assessment by gradient ratioing at visible edges. Image Analysis and Stereology, 2008, 27( 2): 87–95
10	A, Mittal R, Soundararajan A C Bovik . Making a “completely blind” image quality analyzer. IEEE Signal Processing Letters, 2013, 20( 3): 209–212

[1]

FCS-23394-OF-CL_suppl_1

Download

Viewed

Full text

Abstract

Cited

Shared

Discussed