tractatus / bacteria_segmentation

pretrained and instructions for annotating and training segmentation of bacterial fluorescent images

bacteria_segmentation

pretrained and instructions for annotating and training segmentation of bacterial fluorescent images

Installation

Create environment:

conda create --name napari-env
conda activate napari-env
conda install -c conda-forge napari   
conda install opencv

Install Tensorflow for macOS M1/M2:

pip install tensorflow-macos
pip install tensorflow-metal

Install stardist for cell segmentation:

pip install gputools
pip install stardist
pip install csbdeep

pip install splinedist

How to use

Data augmentation

python augment_image_data.py

Perform training

python training.py

Open up tensorboard to follow the results:

tensorboard --logdir=.

Prediction

python prediction.py './examples/' 

Where examples is the folder that contains .tif files for analysis. Format has to be 16-bit monochrome TIFF.

The output will be produced in the active wokring directory as the original file names with prefix output_.

• output_*.tif labels, 16-bit TIFF each cell has a unique unint16 value.
• output_*.zip ImageJ ROI zip file, can be opened in ImageJ/Fiji ROI Manager.