From: Implementing PointNet for point cloud segmentation in the heritage context
Number of building | Types of data | Window (IoU) | Wall (IoU) | Ceiling (IoU) | Door (IoU) | Floor (IoU) | Average (IoU) |
---|---|---|---|---|---|---|---|
Building_25 Building_29 Building_30 | Restitution data | 0.5066 | 0.4588 | 0.8873 | 0.2948 | 0.8211 | 0.4948 |
Building_4_room_19 Building_4_room_20 Building_4_room_21 | 3D Laser Scanner data | 0.0663 | 0.3458 | 0.5481 | 0.8165 | 0.7469 | 0.4206 |
Building_25 Building_29 Building_30 Building_4_room_19 Building_4_room_20 Building_4_room_21 | Restitution Data + 3D Laser Scanner data | 0.4999 | 0.4572 | 0.8782 | 0.3979 | 0.8208 | 0.5090 |