Acta Photonica Sinica, Volume. 51, Issue 3, 0310002(2022)
Mine Dual-band Image Fusion in MS-ADoG Domain Combined with ReNLU and VGG-16
Figures & Tables(10)
Fig. 1. Framework of FIR and VIS image fusion based on Laplace of Gaussian transform
Fig. 4. The functional relationship between the weight w and the critical value T
Table 1. Performance indicators of different algorithms under 4 types of source images
View tableView in ArticleTable 1. Performance indicators of different algorithms under 4 types of source images
Images Indicators Algorithms SVM⁃WLS LatLRR Improved IHS Curvetlet DLF⁃
VGG16
Resnet50⁃
ZCA
Proposed Method EFMMF AG 0.023 0 0.008 3 0.005 9 0.006 6 0.005 5 0.008 0 RMSE 0.175 4 0.161 8 0.319 3 0.160 9 0.160 2 0.222 2 HE 0.490 7 0.497 4 0.502 7 0.499 8 0.499 6 0.504 5 CC 9.735 9 9.655 6 8.787 3 10.167 0 10.303 0 8.467 1 DI 1.424 0 0.451 4 1.502 3 0.552 4 0.560 0 1.148 3 ART 0.910 1 25.795 1 0.859 8 2.855 6 1.797 6 1.692 4 UEC AG 0.039 7 0.013 3 0.008 5 0.013 2 0.010 4 0.0108 RMSE 0.205 3 0.195 0 0.278 1 0.194 5 0.192 5 0.249 2 HE 0.490 4 0.496 1 0.503 7 0.499 6 0.499 6 0.507 0 CC 11.247 7 11.337 8 10.576 0 12.534 3 12.290 2 9.943 5 DI 1.339 4 0.590 7 1.562 8 0.705 6 0.725 4 1.720 5 ART 0.460 7 23.764 7 0.228 3 2.339 7 1.352 4 1.671 8 4⁃3 CATR AG 0.016 1 0.004 8 0.003 9 0.004 2 0.003 4 0.006 0 RMSE 0.233 0 0.214 2 0.309 8 0.208 9 0.209 1 0.290 3 HE 0.485 0 0.495 8 0.506 0 0.499 7 0.499 6 0.512 8 CC 10.970 1 10.890 9 9.789 1 11.796 1 11.770 4 8.982 5 DI 1.398 9 0.532 1 2.801 5 0.606 6 0.602 2 0.832 0 ART 0.448 7 25.951 1 0.205 7 2.410 2 1.235 1 1.708 6 FMMF AG 0.056 1 0.019 1 0.013 6 0.017 9 0.013 6 0.015 3 RMSE 0.190 3 0.168 1 0.259 9 0.161 9 0.162 1 0.207 2 HE 0.494 9 0.497 4 0.502 5 0.499 7 0.499 8 0.503 4 CC 10.557 6 10.144 5 10.649 4 11.283 0 11.244 3 10.267 3 DI 1.612 0 0.370 2 0.503 3 0.418 6 0.427 1 0.834 6 ART 0.452 2 25.113 8 0.225 3 2.996 9 1.285 7 1.660 4
Table 2. Average performance indicators of different fusion methods
View tableView in ArticleTable 2. Average performance indicators of different fusion methods
Fusion algorithms Performance evaluation indicators AG RMSE HE CC DI ART SVM⁃WLS 0.032 7 0.198 7 0.493 0 10.690 9 1.610 0 0.507 0 LatLRR 0.011 5 0.185 3 0.497 3 10.686 7 0.520 0 24.502 4 Improved IHS Curvetlet 0.007 4 0.271 1 0.503 5 10.290 9 1.585 7 0.288 8 DLF⁃VGG16 0.010 4 0.182 4 0.499 7 11.654 1 0.644 2 2.463 5 Resnet50⁃ZCA 0.008 2 0.182 2 0.499 7 11.572 6 0.619 3 1.334 5 Proposed method 0.010 3 0.242 3 0.506 1 9.753 2 1.698 8 1.689 0
Tools
Get Citation
Copy Citation Text
Jiping SUN, Weiqiang FAN. Mine Dual-band Image Fusion in MS-ADoG Domain Combined with ReNLU and VGG-16[J]. Acta Photonica Sinica, 2022, 51(3): 0310002
Paper Information
Category:
Received: Oct. 8, 2021
Accepted: Dec. 9, 2021
Published Online: Apr. 8, 2022
The Author Email: FAN Weiqiang (fan_weiqiang@163.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20