I found skimage's watershed is very slow, So I wrote one here:
https://github.com/Image-Py/imagepy/blob/master/imagepy/ipyalg/hydrology/watershed.py

it is only for 8-bit image(or int image), but I think we can do a histogram equalization, than convert to int image, do watershed.

the performance is OK. a 3d image, which skimage run more than 20 minutes and crash down, my algrthm cost 30s.

----- 原始邮件 -----
发件人:Stefan van der Walt <stefanv@berkeley.edu>
收件人:scikit-image <scikit-image@python.org>
主题:[scikit-image] Comparison of watershed implementations
日期:2018年10月27日 05点48分


I wish the authors also compared accuracy, but here are their
performance benchmarks. No surprises :)
https://www.mdpi.com/2313-433X/4/10/123/pdf
"An Overview of Watershed Algorithm Implementations in Open Source Libraries",
J. Imaging 2018, 4(10), 123; https://doi.org/10.3390/jimaging4100123,
Anton S. Kornilov and Ilia V. Safonov.
Best regards,
Stéfan
_______________________________________________
scikit-image mailing list
scikit-image@python.org
https://mail.python.org/mailman/listinfo/scikit-image