Closing Circle Edges




I need to analyze the size of these particles and managed to get get good segmentation with ‘‘Find Edges’’ after some image processings.
The problem is that when I do ‘‘Analyze Particles’’, many particles are lost since they are not fully-closed circles.

I tried combination of erode and skeletonize to connect the lines but the result is not satisfactory since the segmentation lines are too thin.

Any ideas on how to proceed?




I would advise you to check out the plugin MorphoLibJ - a great tool for segmentation that comes with Fiji*, providing more options/techniques for Segmentation… comes with Morphological Segmentation tool.

I tried using MorphoLibJ’s Morphological Segmentation tool (linked just above) on a cropped section of your image after first applying a Kuwahara filter (sampling window, 5). Here are the settings I used - note: this was just a quick ‘test’ and could very much be refined - and the resulting segmentation looked like:

Perhaps a good option for you?


*NOTE: Fiji is Just ImageJ - it is simply a distribution of ImageJ that comes with a bunch of plugins bundled - ready for you to use out-of-the-box. If you are just getting started, we recommend downloading/using Fiji.



Hi Ellen,

Thank you for your help. I did tried this method before and I noticed some good segmentation but it did not give back a table of the size of the particles such as what ‘‘Analyze Particles’’ gives back when analyzing a binary image.

Is there a way to transform these color particles to distinct particles on a binary image?

Thanks again!



Before you do anything… read through the Morphological Segmentation page on the ImageJ wiki. MorphoLibJ comes with a whole host of measurement tools - but you get different outputs that you can then process to analyze the objects’ sizes, etc.

You can follow this older thread starting here… that should help.