We implement the splitting algorithm for splitting nuclei nucleas described in in 'Nuclei segmentation in histopathology images using deep neural networks'. This algorithm is essentially a dynamic watershed.
The main function is named: post_process.
dynamic_watershed can be installed by unzipping the source code in one directory and using this command: ::
python setup.py install
You can also install it directly from the Python Package Index with this command (not working yet): ::
pip install dynamic_watershed
>>> from dynamic_watershed import post_process
>>> from skimage.io import imread
>>> probability_image = imread('example.png')
>>> p1, p2 = 7, 0.5
>>> result_segmentation = post_process(probability_image, p1, thresh=p2)See file LICENCE.txt in this folder.
dynamic_watershed is an open-source software. Everyone is welcome to contribute !
If you use this work please cite our paper.
BibTeX:
@inproceedings{naylor2017nuclei,
title={Nuclei segmentation in histopathology images using deep neural networks},
author={Naylor, Peter and La{\'e}, Marick and Reyal, Fabien and Walter, Thomas},
booktitle={Biomedical Imaging (ISBI 2017), 2017 IEEE 14th International Symposium on},
pages={933--936},
year={2017},
organization={IEEE}
}