Let me preface that you can do this most easily in ImageJ through their UI.  However, when trying to script this, or actually maniuplate/manage the particles once they've been identified, it's not easy.

So I wrote a library for this purpose in Python.

http://openresearchsoftware.metajnl.com/article/view/jors.bh

Feel free to message me if you want to help getting it to work.   One issue right now is there's a bottleneck that makes it slow on high-res images, so if you have a lot of images, or high-res images, it might be a bit slow.  Also, if you don't need to do much with the particles other than count them, get their basic statistics, you can use scikit-image's labeling API directly.  It's much faster and is why pyparty is leveraging under the hood.

Hi Václav,

If you are comfortable coding in python then give scipy.ndimage module and  scikit-image package a try. Check out section 2.6.6 in the following image processing tutorial using scipy:

http://scipy-lectures.github.io/advanced/image_processing/

You can also check out scikit-image examples for more ideas.

If you are using MATLAB then you can use image-processing toolbox. You may get some ideas/scripts on Matlab central. 

If you are looking for a software to automatically detect and find size of the particles then check for software in medical imaging domain.

- Navtej

I think, the vtk library would fulfil your requests.  It is open source library with python wrapper. 

My group wrote Matlab code to do this some years ago and could probably dig up the code. Message me if it would be helpful.

Hello Václav,

we have developed such algorithms over the last years. If you post some example images, I might be able to give more specific advice

https://opencv.multimedia-processing.com/index.php

check in the website maybe you found dolution

Maybe Matlab can be helpful