Acta Photonica Sinica, Volume. 53, Issue 9, 0910003(2024)
Infrared and Visible Image Fusion Method Based on Information Enhancement and Mask Loss
Figures & Tables(13)
Fig. 3. Comparison diagram before and after source image enhancement
Table 1. Quantitative comparison results of different algorithms on MSRS dataset
View tableView in ArticleTable 1. Quantitative comparison results of different algorithms on MSRS dataset
Methods EN SF AG SD VIF DenseFuse 5.725 4.997 1.635 20.671 0.703 FusionGAN 5.168 3.726 1.219 14.841 0.49 GANMcC 6.05 5.208 1.79 25.765 0.686 RFN-Nest 6.051 5.643 1.87 27.011 0.707 SDNet 4.847 7.023 2.053 14.2 0.486 PIAFusion 6.364 9.549 2.97 38.052 1.002 SeaFusion 6.473 9.294 2.919 37.656 1.016 SwinFusion 6.457 9.315 2.855 38.543 1.03 IRFS 6.369 8.468 2.585 32.327 0.8 Ours 6 . 76411.409 3.54 44.554 1.043
Table 2. Quantitative comparison results of different algorithms on TNO dataset
View tableView in ArticleTable 2. Quantitative comparison results of different algorithms on TNO dataset
Methods EN SF AG SD VIF DenseFuse 6.819 8.985 3.56 34.825 0.658 FusionGAN 6.558 6.275 2.421 30.663 0.422 GANMcC 6.736 6.161 2.544 33.437 0.53 RFN-Nest 6.963 5.874 2.669 36.897 0.559 SDNet 6.695 11.643 4.612 33.669 0.578 PIAFusion 6.814 9.62 3.828 37.141 0.74 SeaFusion 7.133 12.253 4.98 44.244 0.704 SwinFusion 6.891 10.723 4.211 39.447 0.75 IRFS 6.614 8.705 3.167 31.127 0.59 Ours 7.189 15.572 6.321 46.987 0.68
Table 3. Quantitative comparison results of different algorithms on LLVIP dataset
View tableView in ArticleTable 3. Quantitative comparison results of different algorithms on LLVIP dataset
Methods EN SF AG SD VIF DenseFuse 6.597 6.129 1.84 28.19 0.662 FusionGAN 6.288 5.938 1.617 24.341 0.486 GANMcC 6.665 5.728 1.753 31.556 0.625 RFN-Nest 6.833 5.269 1.775 33.719 0.689 SDNet 6.641 9.336 2.59 31.27 0.643 PIAFusion 7.144 10.909 3.102 43.449 0.976 SeaFusion 7.19 10.227 2.877 43.662 0.917 SwinFusion 7.124 10.355 2.91 43.52 0.938 IRFS 6.975 10.101 2.603 38.334 0.834 Ours 7.27 12.528 3.504 53.433 0.998
Table 4. Comparison results of running time of different algorithms （units： s）
View tableView in ArticleTable 4. Comparison results of running time of different algorithms （units： s）
Methods MSRS TNO LLVIP DenseFuse 0.283±0.103 0.607±0.128 0.782±0.158 FusionGAN 0.168±0.019 0.385±0.051 0.462±0.109 GANMcC 0.131±0.027 0.671±0.129 0.711±0.132 RFN-Nest 0.258±0.018 0.214±0.042 0.552±0.127 SDNet 0.038±0.006 0.049±0.018 0.139±0.013 PIAFusion 0.013±0.002 0.028±0.012 0.015±0.004 SeaFusion 0.012±0.001 0.019±0.006 0.012±0.003 SwinFusion 2.258±0.578 3.324±0.878 2.782±0.702 IRFS 0.138±0.006 0.552±0.051 0.236±0.012 Ours 0.057±0.005 0.073±0.013 0.187±0.009
Table 5. Quantitative comparison results of ablation experiments on MSRS dataset 56 pairs of images
View tableView in ArticleTable 5. Quantitative comparison results of ablation experiments on MSRS dataset 56 pairs of images
Stage A B C D EN SF AG SD VIF 1 √ √ √ 6.467 9.353 2.873 38.347 0.986 2 √ √ √ 6.721 11.139 3.461 44.017 1.041 3 √ √ √ 6.618 10.397 3.217 41.421 1.089 4 √ √ √ 6.754 11.315 3.368 43.861 1.01 5 √ √ √ √ 6.764 11.409 3.54 44.554 1.043
Tools
Get Citation
Copy Citation Text
Xiaodong ZHANG, Shuo WANG, Shaoshu GAO, Xinrui WANG, Long ZHANG. Infrared and Visible Image Fusion Method Based on Information Enhancement and Mask Loss[J]. Acta Photonica Sinica, 2024, 53(9): 0910003
Paper Information
Category:
Received: Jan. 29, 2024
Accepted: Apr. 24, 2024
Published Online: Nov. 13, 2024
The Author Email: WANG Shuo (S22070043@s.upc.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20