0

I'm trying to use mesh lab to calculate the relative volumes of point clouds generated from stochastic optical reconstruction microscopy. What that is exactly is not too important. Basically, I have point clouds that represent structures and I need to get some geometric data out of the clouds. Most importantly volume, but other things would be of interest to. The point clouds are mostly roughly spherical/blob shaped. What is the best approach for fitting a mesh to the point clouds to estimate their volumes? Thanks in advance.

Chris Nelson
  • 1
  • 1

1 Answers1

1

If the overall shape of the point cloud is rather convex, a quite robust approach could be to measure just the volume of the convex hull of the point cloud.

Inside MeshLab just type 'convex hull' in the search box. Note that after launching the convex hull filter, there could be need for a coherent reorientation of all the faces (type 'coherenet in the search box) before computing the volume (type volume in the search box and select compute geometric measures)

ALoopingIcon
  • 2,218
  • 1
  • 21
  • 30
  • Thanks, this advice worked well for many of my point clouds. However, occassionally the convex hull is black (instead of the usual grey), and the volume measurement comes back negative. Do you know why this is or how to avoid it? – Chris Nelson Feb 24 '17 at 18:43
  • if the surface is black normals have been oriented inward (wrong, but coherent). So just search and apply the filter `Invert Face Orientation` – ALoopingIcon Feb 24 '17 at 20:58
  • I really appreciate your help. If I might I have one more question: Is there a built in way to selectively export some of the data that has been calculated of an object? Additionally- are there built in macros of any kind? i.e. can I teach meshlab to apply the same functions over and over without having to select them? – Chris Nelson Feb 25 '17 at 21:40