Pixel statistics on labelled images

Authors: Robert Haase, Daniela Vorkel, March 2020

Source

This macro shows how to apply an automated threshold method, how to connect components by labelling, and how to prompt basic measurements using CLIJ.

Get test data:

run("Blobs (25K)");

// open("C:/structure/data/blobs.gif");
getDimensions(width, height, channels, slices, frames);
input = getTitle();

Init GPU and push image data to the GPU

run("CLIJ2 Macro Extensions", "cl_device=");
Ext.CLIJ2_clear();

// push data to GPU
Ext.CLIJ2_push(input);

// clean up ImageJ
run("Close All");

Segment the image and create a label mask

Ext.CLIJ2_thresholdOtsu(input, mask);

// connected components analysis
Ext.CLIJ2_connectedComponentsLabeling(mask, labelmap);

// show labelling
Ext.CLIJx_pull(labelmap);
run("glasbey on dark");

Basic measurements

Based on the labelled mask, we retrieve a table with results.

Ext.CLIJ2_statisticsOfLabelledPixels(input, labelmap);

At the end of the macro, clean up:

Ext.CLIJ2_clear();