loicland / superpoint_graph

Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs
MIT License
758 stars 214 forks source link

how to modify the prune function #198

Closed medif closed 3 years ago

medif commented 4 years ago

Hi, I have a dataset without rgb values with 9 classes labeled and the format is .npy (type float32) I changed the line like: xyz,labels = libply_c.prune(xyz, args.voxel_width,np.zeros(xyz.shape,dtype='u1'),labels,np.zeros(1, dtype='uint8'),9,0) but it throw an error IndexError: vector::_M_range_check: __n (which is 64) >= this->size() (which is 10)

, any suggestions please Thank you

loicland commented 4 years ago

Did you figure it out?

Otherwise please report the value of .shape and dtype for xyz and labels.

medif commented 4 years ago

Hi, It was working fine until i changed my data structure. Now i am using 3 classes (0 as unlabled )

Capture

Can you help me please

loicland commented 4 years ago

why is the label type 'f2'? it should be 'uint8'.