Fine-Grained Point Cloud Semantic Segmentation of Complex Railway Bridge Scenes from UAVs Using Improved DGCNN

<div>Three scene point cloud data synthesized from multiview UAV images. (a) Scene of a simple segmentation task; (b) scene of a moderate segmentation task; (c) scene of a difficult segmentation task.</div>

Structural Control and Health Monitoring

fig2

Figure 2

Figure 2: Fine-Grained Point Cloud Semantic Segmentation of Complex Railway Bridge Scenes from UAVs Using Improved DGCNN 