Machine learning methods to segment cells

I am very new to ML methods. I want to use ML to segment within the pictures the cells and black blobs within them. The problem with simple thresholding is a lot of times the cell outlines where it is darker also gets counted as a black blob. So I was wondering if there is any machine learning algo that I can train to identify the cells and black blobs.

Any direction or guidance will be very much appreciated.

Answer 1

[Shashank Gupta]

Hi,

There are many good segmentation algorithm available which you can try. Its hard to point out one. Check out this file exchange link. It should work for your case, even if it doesn't It will give you a good headstart to explore more. You can try some basic image processing based segmentation and see if the results are apt for your work. Check out this link.

I hope this helps.

Cheers

Answer 2

[Ashish Uthama]

In 23b, you could try using the Cellpose add-on: