Looking in the clustering section but confused

[ML-learner79]

Try to run clustering and getting error not sure why
self.fruit_alpha_shape = self.fruit_alpha_shape_.as_open3d.sample_points_uniformly(1000)
AttributeError: 'GeometryCollection' object has no attribute 'as_open3d'

[ML-learner79]

Furthermore in run_clustering.py
self.fruit_template = o3d.io.read_point_cloud(self.template_path)
print(self.fruit_template.points)
the template is empty.
Not sure that lead to error above

[Cola-1]

Is it figure out now?

[Matt-V50]

Same issue

[sebdeveloper6952]

.ply files are stored using Git-LFS. Make sure to install it and run git lfs pull in the FruitNeRF folder. This will pull the actual contents of the .ply files.

Having said this, I am still running into the error:

AttributeError: 'Trimesh' object has no attribute 'as_open3d'

Reading a bit about Trimesh library, seems like they removed the as_open3d method in some of their recent release. Running pip install trimesh==4.4.0 solves the issue for me.

[Cola-1]

@sebdeveloper6952 Could you please release the newly modified run clustering.py

[Matt-V50]

I believe you need to downgrade trimesh version cause this commit

Try pip install trimesh==4.4.7

[Cola-1]

The final clustering result after modification is still 0.

[EyGy]

The final clustering result after modification is still 0.

Try adjusting the params for downsampling/cleaning
remove_outliers_nb_points and remove_outliers_radius are based on https://www.open3d.org/docs/latest/tutorial/Advanced/pointcloud_outlier_removal.html#Radius-outlier-removal and depending on the scale of your point cloud the default values can lead to removing all of your points. Try adjusting/decreasing the values. Is your result still 0?

[pandachenf]

The final clustering result after modification is still 0.

Hello, have you solved this problem?

[00mmw]

The same issue. Have you solved this problem?

[IceIce1ce]

+1