Point Cloud Support

[baharmon]

A feature request for point cloud support in Dendro for Grasshopper. With point cloud support, Dendro might be able generate volumes from large numbers of points much faster.

[ryein]

thanks for the suggestion. i have this on my list because i think openvdb look ups are quite fast. i have implemented it in custom scripts but havent rolled it into dendro. i havent used it for volume creation though, more for doing operations on large point sets. i am currently trying to address the rendering right now because converting to rhino mesh for previews in each component is a huge performance hit. i will definitely look into after that though. just curious, what would be your typically use case or workflow for point cloud operations? interested to hear your thoughts so as to better understand how best to steer functionality of point cloud features.

…

On Sun, Dec 8, 2024, 2:57 PM Brendan Harmon ***@***.***> wrote: A feature request for point cloud support in Dendro for Grasshopper. With point cloud support, Dendro might be able generate volumes from large numbers of points much faster. — Reply to this email directly, view it on GitHub <#15>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEXJSEDAYEJ7T56WFKUALRL2ETFGBAVCNFSM6AAAAABTHVAWMOVHI2DSMVQWIX3LMV43ASLTON2WKOZSG4ZDKNJYGAZDEMY> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[baharmon]

Glad to hear it is on your list. My typical use for point cloud operations with Dendro would be volumetric modeling for 3D printing. When reconstructing point clouds from airborne lidar, terrestrial laser scanning, or nerfs – especially for scenes with vegetation – I often use Dendro’s Points to Volume operation instead of algorithms like Marching Cubes or Surface Nets. With a high point density, it yields smooth, thin, watertight shells that are good for selective laser sintering or other powder-based printing techniques.

Currently, I start with a large point cloud with millions of points, preprocess and subsample it in CloudCompare (down to 10,000-100,000 points), import it into Rhino, explode it into points, create volumetric model with Dendro’s Points to Volume, optionally apply Boolean operations and smoothing, convert to a mesh, filter out small meshes from outlying points, and then export for 3D printing. Converting a point cloud into points is the major bottleneck as Grasshopper struggles with large numbers of points (100,000+). With a PointCloud to Volume operation for Dendro, I could input much denser point clouds and get higher resolution results with thinner shells.

Other point cloud operations that I’d really like to see implemented include advecting and deforming points for particle simulations.