You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am stuck with the problem of clustering large point clouds into connected segments.
Basically, I want to achieve the results in the Cluster DBSCAN example.
However, my point clouds exceed 10 Mio. Points. and the DBSCAN algorithm becomes really slow in this range.
It would be awsom if there was a way to utilize the open3D Octree functionality to segment large point clouds, but I cannot figure out if or how to code my own method.
Has anyone tried this or dealt with this problem?
The text was updated successfully, but these errors were encountered:
Checklist
main
branch).My Question
Hi Community.
I am stuck with the problem of clustering large point clouds into connected segments.
Basically, I want to achieve the results in the Cluster DBSCAN example.
However, my point clouds exceed 10 Mio. Points. and the DBSCAN algorithm becomes really slow in this range.
The famous CloudCompare software achieves similar Connected Component segmentation with the help of Octree, thus can be more performant.
It would be awsom if there was a way to utilize the open3D Octree functionality to segment large point clouds, but I cannot figure out if or how to code my own method.
Has anyone tried this or dealt with this problem?
The text was updated successfully, but these errors were encountered: