search

charlesq34 / pointnet

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
Other
4.73k stars 1.45k forks source link

dataset preparing #92

Open HannahXu opened 6 years ago

HannahXu commented 6 years ago

I have one or some CAD image(pcl, ply, obj, pcd....) of objects in real word around me ,how can I transfer it to h5 file ? then input this h5 file to pointnet, thanks!

Coldplayplay commented 6 years ago

You can just make your data learning from the style of Stanford3dDataset_v1.2. And then make use of collect_indoor3d_data.py for data re-organization and gen_indoor3d_h5.py to generate HDF5 files, as the author said in Readme.txt.

arunumd commented 4 years ago

You can just make your data learning from the style of Stanford3dDataset_v1.2. And then make use of collect_indoor3d_data.py for data re-organization and gen_indoor3d_h5.py to generate HDF5 files, as the author said in Readme.txt.

I am having issues with accessing pcd files as numpy array for data preparation. I used the pcl_pcd2ply command to first generate a ply file and then used the plyFile API to open the ply file as numpy array. The problem is, the ply file has some duplicate properties and hence wouldn't get parsed in python. Would you be able to suggest an effective way to load pcd files as numpy arrays ?

lyy-1997-hub commented 4 years ago

您可以按照Stanford3dDataset_v1.2的样式进行数据学习。如作者在Readme.txt中所述，然后利用collect_indoor3d_data.py进行数据重组，并使用gen_indoor3d_h5.py生成HDF5文件。 小仙女还在吗，我搞不出来正确的h5文件，想哭