Contour-based meshing of the image domain developed for CINEMAX summer school.
| Input image | Meshing result ael=5 |
Meshing result ael=2 |
|---|---|---|
 |
 |
 |
The module meshing.py contains the functionality for contour-based meshing of the image. Image contours are detected using marching squares implementation from the scikit image package. Computing the conforming constrained Delaunay triangulation relies on the functionality from triangle package, which in turn wraps around Jonathan Richard Shewchuk’s mesh generator.
For 3-phase segmentation check bone meshing notebook, which is also the best documented notebook among the available notebooks.
For 2-phase segmentation check bundles meshing notebook. Chalk and fiber meshing is pretty much the same as for bundles, and bench is an example of meshing an RGB image, by simply converting it to grayscale.