pyntcloud is a Python (3, because we are not in 2008) library for working with 3D point clouds.
📖 📖
| Home |
| Introduction |
| Installation |
| PyntCloud |
| Points |
| Filters // Filters - Dev |
| I/O // I/O - Dev |
| Samplers // Samplers - Dev |
| Structures // Structures - Dev |
You can access most of pyntcloud's functionallity from it's core class: PyntCloud.
With PyntCloud you can penform complex 3D processing operations with minimum lines of code. For example you can:
- Load a point cloud from disk.
- Add 3 new sacalar fields by converting RGB to HSV.
- Build a grid of voxels from the point cloud.
- Build a new point cloud keeping only the nearest point to each occupied voxel center.
- Save the new point cloud.
With the following concise code:
from pyntcloud import PyntCloud
cloud = PyntCloud.from_file("some_file.ply")
cloud.add_scalar_field("hsv")
voxelgrid_id = cloud.add_structure("voxelgrid", x_y_z=[32, 32, 32])
points = cloud.get_sample("voxelgrid_nearest", voxelgrid=voxelgrid_id)
new_cloud = PyntCloud(points)
new_cloud.to_file("out_file.ply")Every PyntCloud can be visualized using the plot method.
This will create a stand-alone html visualizer. The cool thing about this is that you can open it in any browser and if you call it from inside a Jupyter Notebook, the visualizer will be embedded as an IFrame:
The plot function has many options.
For example you can use any scalar field as color with a custom colormap:
Or, if it exists, visualize the mesh associated with the point cloud:
Even though point clouds obtained from different sources present some variance in terms of the kind of information it contains, we encourage a source-agnostic vision of point clouds.
pyntcloud provides tools for source-agnostic 3D processing operations but it also provides building blocks for easily implementing something that covers your specific needs.
Because Python.