Opto-Electronic Engineering, Volume. 50, Issue 4, 220231(2023)
A progressive fusion image enhancement method with parallel hybrid attention
Figures & Tables(16)
Fig. 4. Structure diagram of channel attention and pixel attention
Fig. 6. (a) Progressive feature fusion module；(b) Fusion unit at each stage
Fig. 11. Progressive fusion enhances the effect and details at different stages
Table 1. Compares the amount of LOL data set with advanced image enhancement methods
View tableView in ArticleTable 1. Compares the amount of LOL data set with advanced image enhancement methods
Methods SSIM PSNR /dB LPIPS GMSD FSIM UQI (上箭头 和下箭头 分别表示随着指标数值变大或减小，并将最优结果加粗标出) LIME[ 19 ]0.5649 16.7586 0.4183 0.1541 0.8549 0.8805 MBLLEN[ 20 ]0.7247 17.8583 0.3672 0.1160 0.9262 0.8261 Retinex[ 21 ]0.5997 16.7740 0.4249 0.1549 0.8642 0.9110 KinD[ 22 ]0.8025 20.8741 0.5137 0.0888 0.9397 0.9250 EnGAN [ 23 ]0.6515 17.4828 0.3903 0.1046 0.9226 0.8499 Zero-DCE[ 24 ]0.5623 14.8671 0.3852 0.1646 0.9276 0.7205 GLAD[ 25 ]0.7247 19.7182 0.3994 0.2035 0.9329 0.9204 Ours 0.9053 21.8939 0.3557 0.1035 0.9381 0.9266
Table 2. Compares the values of Lime, DICM and MEF data sets with those of advanced image enhancement methods
View tableView in ArticleTable 2. Compares the values of Lime, DICM and MEF data sets with those of advanced image enhancement methods
Methods LIME DICM MEF LIME[ 30 ]4.1549 3.0005 4.4466 MBLLEN[ 31 ]4.5138 3.6654 4.6901 Retinex[ 32 ]4.5978 4.5779 5.1747 KinD[ 33 ]4.7632 3.5651 4.7514 EnGAN [ 34 ]3.6574 2.9174 3.5373 Zero-DCE[ 35 ]3.7690 2.8348 4.0240 GLAD[ 36 ]4.1282 3.1147 3.6897 Ours 3.4281 2.8054 3.5193
Table 3. Quantitative comparison after adding different network modules to LOL data set
View tableView in ArticleTable 3. Quantitative comparison after adding different network modules to LOL data set
Methods PSNR/dB SSIM Baseline 18.44 0.73 w/o PHA、PFM，with MWA 19.07 0.78 With PHA，w/o MWA、PFM 20.53 0.84 Ours 21.87 0.89
Table 4. Quantitative results of progressive fusion after enhancement in different stages
View tableView in ArticleTable 4. Quantitative results of progressive fusion after enhancement in different stages
Stage PSNR/dB SSIM With 1，w/o 2、3 20.08 0.76 With 1、2，w/o 3 20.91 0.83 With 1、2、3 21.53 0.87
Table 5. Average enhancement time of each method
View tableView in ArticleTable 5. Average enhancement time of each method
Runningtime LIME[ 30 ]GLAD[ 36 ]Enlighten-GAN[ 34 ]KinD[ 33 ]Zero-DCE[ 35 ]Retinex-Net[ 32 ]BIMEF[ 37 ]Ours 20.173 0.0083 0.0053 0.0078 0.0016 0.0063 0.1280 0.0458
Tools
Get Citation
Copy Citation Text
Guanghui Liu, Qi Yang, Yuebo Meng, Minhua Zhao, Hua Yang. A progressive fusion image enhancement method with parallel hybrid attention[J]. Opto-Electronic Engineering, 2023, 50(4): 220231
Paper Information
Category: Article
Received: Sep. 21, 2022
Accepted: Dec. 30, 2022
Published Online: Jun. 15, 2023
The Author Email: Guanghui Liu (guanghuil@163.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20