Dehazing is a wide class of image processing problems, with several categories of solution:
Model-based dehazing uses a model of light propagation into the camera sensor (an Image Formation Model, see [1] and [2]) in conjunction with a statistical prior about the distribution of intensities in images without haze. The success of model-based methods depends significantly on the quality of their priors, which are therefore designed to apply to as restricted a class of images as possible.
Model-free dehazing uses more traditional image processing techniques to improve the quality of hazy images (histogram equalization, color correction, etc.).
Machine learning solutions usually run CNNs on hazy images to predict the ground truth haze-free scene.
Finally, some systems use specialized hardware like polarizing filters, lasers, or multiple camera sensors to gain extra information about a scene before dehazing. I will not consider these.
Thus far, I believe these algorithms represent good starting points to test dehazing performance:
| DCP [3] |
2009 |
Model-based |
Air |
Image |
| UDCP [4] |
2013 |
Model-based |
Water |
Image |
| MIP [5] |
2010 |
Model-based |
Water |
Image |
| HL [6] |
2016 |
Model-based |
Air |
Image |
| HL (underwater) [7] |
2021 |
Model-based |
Water |
Image |
| ULAP [8] |
2018 |
Model-based |
Water |
Image |
| DCP + TF [9] |
2016 |
Model-based |
Air |
Video |
| UWCNN [10] |
2020 |
Learning |
Water |
Video |
| WaterNet [11] |
2020 |
Learning |
Water |
Image |
| UIEC^2-Net [12] |
2021 |
Learning |
Water |
Image |
Some of these models were selected for their high performance and recent publication, but some were selected because they represent baseline algorithms against which to compare other models. I still need to identify pipelines that fit these criteria in the model-free domain, and to find more recent algorithms which target video dehazing.
[1]
J. S. Jaffe,
“Computer modeling and the design of optimal underwater imaging systems,” IEEE Journal of Oceanic Engineering, vol. 15, no. 2, pp. 101–111, Apr. 1990, doi:
10.1109/48.50695.
[2]
Y. Y. Schechner and N. Karpel,
“Clear underwater vision,” in
Proceedings of the 2004 IEEE computer society conference on computer vision and pattern recognition, 2004. CVPR 2004., Jun. 2004, pp. I–I. doi:
10.1109/CVPR.2004.1315078.
[3]
K. He, J. Sun, and X. Tang,
“Single image haze removal using dark channel prior,” in
2009 IEEE conference on computer vision and pattern recognition, Jun. 2009, pp. 1956–1963. doi:
10.1109/CVPR.2009.5206515.
[4]
P. Drews, E. R. Nascimento, F. Moraes, S. S. C. Botelho, and M. F. Montenegro Campos,
“Transmission estimation in underwater single images,” in
2013 IEEE international conference on computer vision workshops, Dec. 2013, pp. 825–830. doi:
10.1109/ICCVW.2013.113.
[5]
N. Carlevaris-Bianco, A. Mohan, and R. M. Eustice,
“Initial results in underwater single image dehazing,” in
OCEANS 2010 MTS/IEEE SEATTLE, Sep. 2010, pp. 1–8. doi:
10.1109/OCEANS.2010.5664428.
[6]
D. Berman, T. Treibitz, and S. Avidan,
“Non-local image dehazing,” in
2016 IEEE conference on computer vision and pattern recognition (CVPR), Jun. 2016, pp. 1674–1682. doi:
10.1109/CVPR.2016.185.
[7]
D. Berman, D. Levy, S. Avidan, and T. Treibitz,
“Underwater single image color restoration using haze-lines and a new quantitative dataset,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 43, no. 8, pp. 2822–2837, Aug. 2021, doi:
10.1109/TPAMI.2020.2977624.
[8]
W. Song, Y. Wang, D. Huang, and D. Tjondronegoro, “A rapid scene depth estimation model based on underwater light attenuation prior for underwater image restoration,” in Advances in multimedia information processing – PCM 2018, R. Hong, W.-H. Cheng, T. Yamasaki, M. Wang, and C.-W. Ngo, Eds., Cham: Springer International Publishing, 2018, pp. 678–688.
[9]
C. Qing, F. Yu, X. Xu, W. Huang, and J. Jin,
“Underwater video dehazing based on spatial–temporal information fusion,” Multidim Syst Sign Process, vol. 27, no. 4, pp. 909–924, Oct. 2016, doi:
10.1007/s11045-016-0407-2.
[10]
C. Li, S. Anwar, and F. Porikli,
“Underwater scene prior inspired deep underwater image and video enhancement,” Pattern Recognition, vol. 98, p. 107038, Feb. 2020, doi:
10.1016/j.patcog.2019.107038.
[11]
C. Li
et al.,
“An underwater image enhancement benchmark dataset and beyond,” IEEE Transactions on Image Processing, vol. 29, pp. 4376–4389, 2020, doi:
10.1109/TIP.2019.2955241.
[12]
Y. Wang, J. Guo, H. Gao, and H. Yue,
“UIEC^2-net: CNN-based underwater image enhancement using two color space,” Signal Processing: Image Communication, vol. 96, p. 116250, Aug. 2021, doi:
10.1016/j.image.2021.116250.