DefogNet: A Single-Image Dehazing Algorithm with Cyclic Structure and Cross-Layer Connections
Inspired by the application of CycleGAN networks to the image style conversion problem Zhu et al. (2017), this paper proposes an end-to-end network, DefogNet, for solving the single-image dehazing problem, treating the image dehazing problem as a style conversion problem from a fogged image to a non...
| Auteurs principaux: | , , , , , |
|---|---|
| Format: | Article |
| Langue: | English |
| Publié: |
Hindawi-Wiley
2021-01-01
|
| Collection: | Complexity |
| Accès en ligne: | http://dx.doi.org/10.1155/2021/2352185 |
| _version_ | 1827078928593846272 |
|---|---|
| author | Suting Chen Wenhao Fan Shaw Peter Chuang Zhang Kui Chen Yong Huang |
| author_facet | Suting Chen Wenhao Fan Shaw Peter Chuang Zhang Kui Chen Yong Huang |
| author_sort | Suting Chen |
| collection | DOAJ |
| description | Inspired by the application of CycleGAN networks to the image style conversion problem Zhu et al. (2017), this paper proposes an end-to-end network, DefogNet, for solving the single-image dehazing problem, treating the image dehazing problem as a style conversion problem from a fogged image to a nonfogged image, without the need to estimate a priori information from an atmospheric scattering model. DefogNet improves on CycleGAN by adding a cross-layer connection structure in the generator to enhance the network’s multiscale feature extraction capability. The loss function was redesigned to add detail perception loss and color perception loss to improve the quality of texture information recovery and produce better fog-free images. In this paper, the novel Defog-SN algorithm is presented. This algorithm adds a spectral normalization layer to the discriminator’s convolution layer to make the discriminant network conform to a 1-Lipschitz continuum and further improve the model’s stability. In this study, the experimental process is completed based on the O-HAZE, I-HAZE, and RESIDE datasets. The dehazing results show that the method outperforms traditional methods in terms of PSNR and SSIM on synthetic datasets and Avegrad and Entropy on naturalistic images. |
| first_indexed | 2025-03-20T02:37:28Z |
| format | Article |
| id | doaj.art-e6db768d4dec41dbb5ff7bcc6a9a2022 |
| institution | Directory Open Access Journal |
| issn | 1076-2787 1099-0526 |
| language | English |
| last_indexed | 2025-03-20T02:37:28Z |
| publishDate | 2021-01-01 |
| publisher | Hindawi-Wiley |
| record_format | Article |
| series | Complexity |
| spelling | doaj.art-e6db768d4dec41dbb5ff7bcc6a9a20222024-10-03T07:43:41ZengHindawi-WileyComplexity1076-27871099-05262021-01-01202110.1155/2021/23521852352185DefogNet: A Single-Image Dehazing Algorithm with Cyclic Structure and Cross-Layer ConnectionsSuting Chen0Wenhao Fan1Shaw Peter2Chuang Zhang3Kui Chen4Yong Huang5Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science & Technology, Nanjing 210044, ChinaJiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science & Technology, Nanjing 210044, ChinaJiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science & Technology, Nanjing 210044, ChinaJiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science & Technology, Nanjing 210044, ChinaJiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science & Technology, Nanjing 210044, ChinaThe Anhui Province Meteorological Science Research Institute, Hefei 230061, ChinaInspired by the application of CycleGAN networks to the image style conversion problem Zhu et al. (2017), this paper proposes an end-to-end network, DefogNet, for solving the single-image dehazing problem, treating the image dehazing problem as a style conversion problem from a fogged image to a nonfogged image, without the need to estimate a priori information from an atmospheric scattering model. DefogNet improves on CycleGAN by adding a cross-layer connection structure in the generator to enhance the network’s multiscale feature extraction capability. The loss function was redesigned to add detail perception loss and color perception loss to improve the quality of texture information recovery and produce better fog-free images. In this paper, the novel Defog-SN algorithm is presented. This algorithm adds a spectral normalization layer to the discriminator’s convolution layer to make the discriminant network conform to a 1-Lipschitz continuum and further improve the model’s stability. In this study, the experimental process is completed based on the O-HAZE, I-HAZE, and RESIDE datasets. The dehazing results show that the method outperforms traditional methods in terms of PSNR and SSIM on synthetic datasets and Avegrad and Entropy on naturalistic images.http://dx.doi.org/10.1155/2021/2352185 |
| spellingShingle | Suting Chen Wenhao Fan Shaw Peter Chuang Zhang Kui Chen Yong Huang DefogNet: A Single-Image Dehazing Algorithm with Cyclic Structure and Cross-Layer Connections Complexity |
| title | DefogNet: A Single-Image Dehazing Algorithm with Cyclic Structure and Cross-Layer Connections |
| title_full | DefogNet: A Single-Image Dehazing Algorithm with Cyclic Structure and Cross-Layer Connections |
| title_fullStr | DefogNet: A Single-Image Dehazing Algorithm with Cyclic Structure and Cross-Layer Connections |
| title_full_unstemmed | DefogNet: A Single-Image Dehazing Algorithm with Cyclic Structure and Cross-Layer Connections |
| title_short | DefogNet: A Single-Image Dehazing Algorithm with Cyclic Structure and Cross-Layer Connections |
| title_sort | defognet a single image dehazing algorithm with cyclic structure and cross layer connections |
| url | http://dx.doi.org/10.1155/2021/2352185 |
| work_keys_str_mv | AT sutingchen defognetasingleimagedehazingalgorithmwithcyclicstructureandcrosslayerconnections AT wenhaofan defognetasingleimagedehazingalgorithmwithcyclicstructureandcrosslayerconnections AT shawpeter defognetasingleimagedehazingalgorithmwithcyclicstructureandcrosslayerconnections AT chuangzhang defognetasingleimagedehazingalgorithmwithcyclicstructureandcrosslayerconnections AT kuichen defognetasingleimagedehazingalgorithmwithcyclicstructureandcrosslayerconnections AT yonghuang defognetasingleimagedehazingalgorithmwithcyclicstructureandcrosslayerconnections |