pyimagej.md

Running ImageJ macros from python: pyimagej

There is a huge community of image data scientists used to ImageJ macro and also a large community who use python libraries such as scikit-image and scipy to process and analyse their image data. Obviously, both communities enjoy certain advantages of their ecosystem. Goal of the pyimagej project is to build bridges between these two ecosystems and in particular enable the python community to run ImageJ plugins within their python workflows. Before we can get started, we should download and install conda. Then, we can setup a conda environment specifically for pyimagej as explained in pyimagej's installation instructions:

conda config --add channels conda-forge
conda config --set channel_priority strict

conda create -n pyimagej pyimagej openjdk=8

Just to explain this command a bit. It mentions "pyimagej" twice, intentionally. The first one comes with a decorator: -n pyimagej. The -n is a short version of --name and declares that the conda environment we create will be called "pyimagej". The second "pyimagej" specifies a conda package named pyimagej. Thus, we are installing pyimagej into a conda environment named "pyimagej". Furthermore, we also install a Java Development Kit (JDK), version 8, into this conda environment as specified with openjdk=8.

This will install some packages:

We then just do what the terminal suggests us to do, we activate the environment, again named "pyimagej":

conda activate pyimagej

In order to test pyimagej, we also need to install some more stuff such as jupyter notebooks and scikit-image for loading and visualizing images:

conda install jupyter notebook scikit-image

We can then test the installation, e.g. by executing the following code in a juypter notebook. Therefore, we start jupyter by entering this on the command line:

jupyter notebook pyimagej_first_test.ipynb

Afterwards, we paste this code in the notebook and enter SHIFT-Enter

import imagej
import os
from skimage.io import imread, imshow, imsave

# determine the current folder; and deal with backslashes \ on Windows; all operating systems acceppt /
current_directory = os.getcwd().replace("\\", "/") + "/"

# Start up ImageJ
ij = imagej.init()

# define input and output
source_image_file = 'https://wsr.imagej.net/images/Cell_Colony.jpg'
result_image_file = current_directory + "test.tif"

# define code to run
macro_code = """
open("{source_image_file}");
run("Find Edges");
saveAs("Tiff", "{result_image_file}");
""".format(
    source_image_file = source_image_file, 
    result_image_file = result_image_file
)

# execute the macro
ij.py.run_macro(macro_code)

# show input and output
source_image = imread(source_image_file)
imshow(source_image)

result_image = imread(result_image_file)
imshow(result_image)

... that should result in this output: