3D EDT can be much quicker

seung-lab / euclidean-distance-transform-3d

Euclidean distance & signed distance transform for multi-label 3D anisotropic images using marching parabolas.

GNU General Public License v3.0

234 stars 37 forks source link

Closed Pavlik1400 closed 6 months ago

Pavlik1400 commented 6 months ago

I have noticed that parallelization implementation for three-dimensional EDT is not implemented efficiently and can be significantly accelerated. Please check PR https://github.com/seung-lab/euclidean-distance-transform-3d/pull/50

william-silversmith commented 6 months ago

Resolved in #50

Thank you! This will be released in PyPI version 2.4.0