Fast Mixture Model Segmentation

Fast mixture model segmentation used in Boone and Andrews labs

Python requirements

• Numpy
• Cython
• scikit-image

Installation

Create a virtual environment (optional)

$ virtualenv -ppython3 segmentation-env
$ source segmentation-env/bin/activate

Install python requirements (needed to build the package)

$ pip install numpy cython

Install our library (pulls in all other dependencies)

$ pip install segmentation

Usage example

import segmentation as seg
from skimage.io import imread

image = imread('./001001000.tiff', plugin='tifffile')[1]  # Read channel 1 of a tiff/flex 
im = seg.blur_frame(image) # gaussian blur
segmented, _ = seg.mixture_model(im, debug=True) # second return argument is currently unused
labels = seg.watershed(im, segmented)