| 模型 | ACC | Precision | Recall |
| ResNet101 [27] | 0.920 | 0.915 | 0.915 |
| ResNet34 [27] | 0.910 | 0.885 | 0.905 |
| VGG_16 [28] | 0.900 | 0.900 | 0.890 |
| VGG_19 [28] | 0.870 | 0.855 | 0.865 |
| GoogLeNet [29] | 0.910 | 0.900 | 0.900 |
| SqueezeNet [30] | 0.900 | 0.900 | 0.865 |
| DenseNet [31] | 0.880 | 0.855 | 0.890 |
| EfficientNet [32] | 0.851 | 0.852 | 0.799 |
| SwinTransformer [33] | 0.795 | 0.778 | 0.732 |
| ViT [34] | 0.821 | 0.857 | 0.735 |
| 本文模型 | 0.966 | 0.968 | 0.980 |