Research Article
SACTNet: Spatial Attention Context Transformation Network for Cloud Removal
Table 3
Comparison of cloud removal effects on RICE1 and RICE2 images.
| | RICE1 | RICE2 | PSNR | SSIM | RMSE | PSNR | SSIM | RMSE |
| Cloud-GAN [23] | 27.656 | 0.935 | 4.124 | 26.5385 | 0.8758 | 4.363 | CGAN [25] | 26.547 | 0.903 | 5.323 | 25.384 | 0.811 | 5.213 | CycleGAN [26] | 25.88 | 0.893 | 5.121 | 23.91 | 0.793 | 6.091 | SpA-GAN [27] | 30.232 | 0.954 | 4.765 | 28.368 | 0.906 | 5.857 | RSCNet [28] | 28.112 | 0.951 | 3.212 | 26.168 | 0.865 | 5.451 | GCANet [29] | 29.568 | 0.961 | 3.583 | 27.322 | 0.916 | 6.201 | MLA-GAN [30] | 31.858 | 0.971 | 4.031 | 31.124 | 0.938 | 4.391 | Our results | 33.251 | 0.989 | 5.335 | 32.851 | 0.958 | 6.338 |
|
|