Attention mechanism enhancement algorithm based on cycle consistent generative adversarial networks for single image dehazing

被引:12
作者
Liu, Yan [1 ]
Al-Shehari, Hassan [1 ]
Zhang, Hongying [2 ]
机构
[1] Southwest Jiaotong Univ, Sch Comp & Artificial Intelligence, Chengdu 611756, Sichuan, Peoples R China
[2] Southwest Univ Sci & Technol, Robot Technol Used Special Environm Key Lab, Mianyang 621010, Sichuan, Peoples R China
关键词
Image dehazing; Attention mechanism; Deep learning; Generative adversarial networks;
D O I
10.1016/j.jvcir.2021.103434
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
This paper proposes AMEA-GAN, an attention mechanism enhancement algorithm. It is cycle consistency-based generative adversarial networks for single image dehazing, which follows the mechanism of the human retina and to a great extent guarantees the color authenticity of enhanced images. To address the color distortion and fog artifacts in real-world images caused by most image dehazing methods, we refer to the human visual neurons and use the attention mechanism of similar Horizontal cell and Amazon cell in the retina to improve the structure of the generator adversarial networks. By introducing our proposed attention mechanism, the effect of haze removal becomes more natural without leaving any artifacts, especially in the dense fog area. We also use an improved symmetrical structure of FUNIE-GAN to improve the visual color perception or the color authenticity of the enhanced image and to produce a better visual effect. Experimental results show that our proposed model generates satisfactory results, that is, the output image of AMEA-GAN bears a strong sense of reality. Compared with state-of-the-art methods, AMEA-GAN not only dehazes images taken in daytime scenes but also can enhance images taken in nighttime scenes and even optical remote sensing imagery.
引用
收藏
页数:11
相关论文
共 59 条
[1]   Day and Night-Time Dehazing by Local Airlight Estimation [J].
Ancuti, Cosmin ;
Ancuti, Codruta O. ;
De Vleeschouwer, Christophe ;
Bovik, Alan C. .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2020, 29 :6264-6275
[2]  
Ancuti C, 2016, IEEE IMAGE PROC, P2256, DOI 10.1109/ICIP.2016.7532760
[3]  
[Anonymous], 2019, ARXIV190100600V1
[4]  
Arjovsky M., 2014, ARXIV EPRINTS
[5]   Single Image Dehazing Using Haze-Lines [J].
Berman, Dana ;
Treibitz, Tali ;
Avidan, Shai .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2020, 42 (03) :720-734
[6]   Non-Local Image Dehazing [J].
Berman, Dana ;
Treibitz, Tali ;
Avidan, Shai .
2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2016, :1674-1682
[7]   DehazeNet: An End-to-End System for Single Image Haze Removal [J].
Cai, Bolun ;
Xu, Xiangmin ;
Jia, Kui ;
Qing, Chunmei ;
Tao, Dacheng .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2016, 25 (11) :5187-5198
[8]   Gated Context Aggregation Network for Image Dehazing and Deraining [J].
Chen, Dongdong ;
He, Mingming ;
Fan, Qingnan ;
Liao, Jing ;
Zhang, Liheng ;
Hou, Dongdong ;
Yuan, Lu ;
Hua, Gang .
2019 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV), 2019, :1375-1383
[9]  
Chen ZH, 2018, 2018 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), P2626, DOI 10.1109/ICASSP.2018.8462078
[10]   Deep Multi-Model Fusion for Single-Image Dehazing [J].
Deng, Zijun ;
Zhu, Lei ;
Hu, Xiaowei ;
Fu, Chi-Wing ;
Xu, Xuemiao ;
Zhang, Qing ;
Qin, Jing ;
Heng, Pheng-Ann .
2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2019), 2019, :2453-2462