kaolin.conversions.sdfconversions

sdf_to_pointcloud(sdf: <module 'kaolin.rep.SDF' from '/home/docs/checkouts/readthedocs.org/user_builds/kaolin-jfl/envs/latest/lib/python3.6/site-packages/kaolin-0.2.0+756dba9-py3.6-linux-x86_64.egg/kaolin/rep/SDF.py'>, bbox_center: float = 0.0, bbox_dim: float = 1.0, resolution: int = 32, upsampling_steps: int = 2, num_points: int = 5000)[source]

Converts an SDF fucntion to a point cloud.

Parameters

  • sdf (kaolin.rep.SDF) – an object with a .eval_occ function that indicates which of a set of passed points is inside the surface.

  • bbox_center (float) – center of the surface’s bounding box.

  • bbox_dim (float) – largest dimension of the surface’s bounding box.

  • resolution (int) – the initial resolution of the voxel, should be large enough to properly define the surface.

  • upsampling_steps (int) – Number of times the initial resolution will be doubled. The returned resolution will be resolution * (2 ^ upsampling_steps)

  • num_points (int) – number of points in computed point cloud.

Returns

computed point cloud

Return type

(torch.FloatTensor)

Example

>>> sdf = kal.rep.SDF.sphere()
>>> points = kal.conversion.sdf_to_pointcloud(sdf, bbox_dim=2)
sdf_to_trianglemesh(sdf: <module 'kaolin.rep.SDF' from '/home/docs/checkouts/readthedocs.org/user_builds/kaolin-jfl/envs/latest/lib/python3.6/site-packages/kaolin-0.2.0+756dba9-py3.6-linux-x86_64.egg/kaolin/rep/SDF.py'>, bbox_center: float = 0.0, bbox_dim: float = 1.0, resolution: int = 32, upsampling_steps: int = 2)[source]

Converts an SDF function to a mesh

Parameters

  • sdf (kaolin.rep.SDF) – an object with a .eval_occ function that indicates which of a set of passed points is inside the surface.

  • bbox_center (float) – center of the surface’s bounding box.

  • bbox_dim (float) – largest dimension of the surface’s bounding box.

  • resolution (int) – the initial resolution of the voxel, should be large enough to properly define the surface.

  • upsampling_steps (int) – Number of times the initial resolution will be doubled. The returned resolution will be resolution * (2 ^ upsampling_steps)

Returns

computed mesh preperties

Return type

(torch.Tensor)

Example

>>> sdf = kal.rep.SDF.sphere()
>>> verts, faces = kal.conversion.sdf_to_trianglemesh(sdf, bbox_dim=2)
>>> mesh = kal.rep.TriangleMesh.from_tensors(verts, faces)
sdf_to_voxelgrid(sdf: <module 'kaolin.rep.SDF' from '/home/docs/checkouts/readthedocs.org/user_builds/kaolin-jfl/envs/latest/lib/python3.6/site-packages/kaolin-0.2.0+756dba9-py3.6-linux-x86_64.egg/kaolin/rep/SDF.py'>, bbox_center: float = 0.0, bbox_dim: float = 1.0, resolution: int = 32, upsampling_steps: int = 2)[source]

Converts an SDF to a voxel grid.

Parameters

  • sdf (kaolin.rep.SDF) – an object with a .eval_occ function that indicates which of a set of passed points is inside the surface.

  • bbox_center (float) – center of the surface’s bounding box.

  • bbox_dim (float) – largest dimension of the surface’s bounding box.

  • resolution (int) – the initial resolution of the voxel, should be large enough to properly define the surface.

  • upsampling_steps (int) – Number of times the initial resolution will be doubled. The returned resolution will be resolution * (2 ^ upsampling_steps)

Returns

a voxel grid

Return type

(torch.Tensor)

Example

>>> sdf = kal.rep.SDF.sphere()
>>> voxel = kal.conversions.sdf_to_voxelgrid(sdf, bbox_dim = 2)