Use SurfaceNets for surface generation #1
Description
The current implementation use marching cubes to extract the surfaces from the segmented image. Surface nets could be both fast and more accurate (especially for touching labels).
Links:
https://docs.pyvista.org/version/stable/api/core/_autosummary/pyvista.ImageDataFilters.contour_labeled.html
https://jcgt.org/published/0011/01/03/paper.pdf