I have a pointcloud which I imported to the pyntcloud libray as a series of points, it is a fully 3D pointcloud, as in it bounds forms a volume.
points = pd.DataFrame(points)
points.columns = ['x', 'y', 'z']
cloud = PyntCloud(points)
I calculate the normals
k_neighbors = cloud.get_neighbors(k=10)
cloud_norm=cloud
cloud_norm.add_scalar_field("normals", k_neighbors=k_neighbors)
I would like to generate a solid object, preferably a polyhedral, I've looked at the CGAL bindings and pymesh but I'm not finding a working solution. Any ideas?
The data is structured like this
cloud.points
x y z nx(11) ny(11) nz(11)
0 991.538025 267.574707 191.911194 -0.065332 -0.106776 0.992134
1 991.545227 267.598602 191.912704 -0.157886 0.069813 0.984986
2 991.546570 267.587189 191.913498 -0.124825 -0.085891 0.988454
3 991.548889 267.565887 191.910797 -0.082405 -0.210254 0.974168
4 991.549805 267.563507 191.911499 -0.094382 -0.256764 0.961855
5 991.552124 267.624298 191.910599 -0.192515 0.238861 0.951779
CloudCompare has a plugin that can do this. It uses an open source plugin that does "Poisson reconstruction", see https://www.cs.jhu.edu/~misha/Code/PoissonRecon/Version12.00/ It Generates .ply files.