How to label point clouds for semantic segmentation?

[TuanNguyenSKKU]

Hi everyone,

Can you suggest me an effective tool to label point clouds for semantic segmentation? I have already had a point cloud file with 6 columns (XYZ & RGB), and now I want to add one last column as label.

Thank you very much.

[bhagatdas]

You don't need to add the Label Mannually.. the indoor3d_util.py will add the labels for you. see.. points = np.loadtxt(f) labels = np.ones((points.shape[0],1)) * g_class2label[cls] // 6 featue to 7 feature points_list.append(np.concatenate([points, labels], 1)) # Nx7

[wangsale]

CloudCompare

[SiddhikaArun]

You don't need to add the Label Mannually.. the indoor3d_util.py will add the labels for you. see.. points = np.loadtxt(f) labels = np.ones((points.shape[0],1)) * g_class2label[cls] // 6 featue to 7 feature points_list.append(np.concatenate([points, labels], 1)) # Nx7

Hi, did you try implementing PointNet++ on own dataset?