基于不同深度学习模型提取建筑轮廓的方法研究
|
|
胡腾云,解鹏飞,温亚楠,慕号伟
|
Research on Building Footprints Extraction Methods based on Different Deep Learning Models
|
|
Tengyun HU,Pengfei XIE,Yanan WEN,Haowei MU
|
|
表2 不同深度学习模型精度对比
|
Table 2 Accuracy comparison of different deep learning models
|
|
| 骨干网络（backbone） | 精确率（Precision） | 召回率（Recall） | F1指数 | IoU | 训练时长/h |
|---|
| U-Net | -- | 88.98% | 87.84% | 88.09% | 79.37% | 20 | | DANet | -- | 81.68% | 76.58% | 78.46% | 65.59% | 30 | | UA-Net | -- | 83.82% | 82.01% | 82.30% | 71.03% | 35 | | Mask R-CNN | ResNet | 80.62% | 72.46% | 75.34% | 61.82% | 17 | | Mask R-CNN FPN | ResNet FPN | 83.13% | 66.65% | 67.97% | 52.53% | 31 | | Mask R-CNN RX FPN | ResNeXt FPN | 85.14% | 66.36% | 73.61% | 59.70% | 43 |
|
|
|