Detecting Circular Patterns in Microscopy Image


I am fairly new to ImageJ and want to use it to detect/measure particles (or fibres in that case) in an Image obtained from optical microscopy.
The goal is to measure the area of the particles relative to the total area. Counting particles would be a start, tough.
This is the image I am starting with:

I applied these steps using Edge-Detection and Hough Circles.

run("Enhance Contrast...", "saturated=0.3 equalize");
run("FeatureJ Edges", "compute smoothing=4 suppress lower=[] higher=4");
run("Hough Circles", "minimum=10 maximum=20 increment=2 number=100 threshold=60");

It detects or rather finds the centres of the fibres as supposed, but it takes long to compute (if I want to find all the particles) and detects too many patricles if the number is increased. However here is a result:

The “standard” Analyze Patricles function does not seem to fit, as it looks for threshholded patricles only, while mine are only to be recognized by their grey surrouding…

Is there an easier/faster way of doing this?

Thank you for your help in advance.


The simplest I found is using a Laplacian Filter on your data, followed by find Maxima

From the image you submitted

run("FeatureJ Laplacian", "compute smoothing=4");
run("Find Maxima...", "noise=1.50 output=[Point Selection] light");

The Laplacian filter that I used is part of FeatureJ available from the great ImageScience update site.

1 Like