is_manifold vs is_closed

[danafd]

Hello,
I'm fairly new to analysing meshes and sometimes get confused with terminologies that I am not familiar with. What is the difference between the is_manifold() and is_closed() functions for meshes? And which one do I need to be true to measure the true volume of a mesh ( using the volume() function)?

[marcomusy]

Hi, for a mesh to be manifold it must not have no edge with more than 2 faces attached to it.
Try

>>> from vedo import *
>>> Circle().is_manifold()
True
>>> Circle().is_closed()
False
>>> Sphere().is_closed()
True
>>> Sphere().is_manifold()
True

[danafd]

Thanks for the clarification @marcomusy!
I have the following 3D shape (binary segmentation from microscopy image of embryonic forebrain). I would like to quantify its volume but I'm wary of the possibility that the volume() function would give false results because the mesh is not closed. I show the image at an angle where it is clear that the 2 big lobes are not connected in space.

msh.is_manfold() returns True
msh.is_closed() returns False

Does the msh.volume() work here?

Thanks for your help!

[marcomusy]

Forn non closed meh the result may be not precise.
I would check where it is not closed with mesh.boundaries().
you can also use mesh.split() to split in 2 (or more pieces) and evaluate the volumes separately.
you can also use .extract_largest_region() to remove the noise.

[danafd]

Thanks for the tips!