全局信息提取与重建的遥感图像语义分割网络
|
|
梁龙学,贺成龙,吴小所,闫浩文
|
Remote sensing image semantic segmentation network based on global information extraction and reconstruction
|
|
Longxue LIANG,Chenglong HE,Xiaosuo WU,Haowen YAN
|
|
| 表 2 在Potsdam测试集与先进的遥感语义分割网络结果进行对比 |
| Tab.2 Comparison results on Potsdam test set with state-of-art remote sensing semantic segmentation network |
|
| 方法 | 骨干 | C/MB | Np/106 | Nf/109 | F1mean/% | OA/% | mIoU/% | | DANet(2019) | Resnet18 | 2024.9 | 12.6 | 120.24 | 90.7 | 90.2 | 83.1 | | BANet(2021) | ResTLi | 3248.0 | 12.7 | 29.38 | 92.1 | 90.6 | 85.6 | | ABCNet(2021) | Resnet18 | 1573.2 | 14.0 | 62.16 | 92.2 | 90.8 | 85.8 | | MANet(2021) | Resnet18 | 2091.6 | 12.0 | 88.25 | 92.4 | 90.8 | 86.1 | | UnetFormer(2022) | Resnet18 | 1481.7 | 11.7 | 11.67 | 92.7 | 91.1 | 86.6 | | MAResUNet(2022) | Resnet18 | 638.51 | 16.2 | 25.29 | 92.7 | 91.3 | 86.7 | | DCswin(2022) | Swin-tiny | 4265.9 | 45.6 | 89.30 | 92.9 | 91.3 | 86.9 | | MAGIFormer | MSCAN_tiny | 5015.3 | 13.9 | 62.70 | 93.0 | 91.4 | 87.1 |
|
|
|