https://github.com/mzur/watershed

A Python implementation of the watershed image segmentation algorithm
https://github.com/mzur/watershed

image-processing python watershed-algorithm

Last synced: about 2 months ago
JSON representation

A Python implementation of the watershed image segmentation algorithm

• Host: GitHub
• URL: https://github.com/mzur/watershed
• Owner: mzur
• License: mit
• Created: 2017-10-24T14:06:09.000Z (about 7 years ago)
• Default Branch: master
• Last Pushed: 2017-10-25T09:15:56.000Z (about 7 years ago)
• Last Synced: 2024-05-02T01:10:56.532Z (8 months ago)
• Topics: image-processing, python, watershed-algorithm
• Language: Python
• Size: 12.7 KB
• Stars: 75
• Watchers: 2
• Forks: 34
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Watershed

A simple (but not very fast) Python implementation of [Determining watersheds in digital pictures via flooding simulations](http://dx.doi.org/10.1117/12.24211).

![source image](ex.png) ![labelled image](ws.png)

In contrast to [`skimage.morphology.watershed`](http://scikit-image.org/docs/dev/api/skimage.morphology.html#skimage.morphology.watershed) and [`cv2.watershed`](https://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_imgproc/py_watershed/py_watershed.html) this implementation does not use marker seeds.

## Usage

```python

import numpy as np

from Watershed import Watershed

from PIL import Image

import matplotlib.pyplot as plt

w = Watershed()

image = np.array(Image.open('ex.png'))

labels = w.apply(image)

plt.imshow(labels, cmap='Paired', interpolation='nearest')

plt.show()

```