Estimating thickness of filaments in EM images using Ridge Detection

Hello everyone,

I have some images with filamentous structures I want to estimate the width of. Can Ridge Detection plugin be used for this? I am concerned because there is a mandatory parameter called sigma which is associated with the width that user has to specify.
Or is the ‘estimate width’ option of Ridge Detection plugin already good enough solution to have an unbiased width measurement?

Thanks in advance!


Yes. :slight_smile: You can search here on the forum for posts that might be relevant in this case. One recent one might be: How to measure fiber length and diameter automatically?

I don’t know. Perhaps @twagner is the best person to answer this question… unless someone else here knows?

1 Like

The estimate width is good enough. However , the upper and lower threshold is more like a good guess and often has to be optimized.


Thanks a lot @twagner @etadobson !
Could you tell me how the user defining mandatory parameter sigma value affect the width estimation?
Doesn’t it mean we are somehow forcing it to give a width based on the linear relation of sigma and w?

I could be missing something.

Btw, @twagner, I would like to put a word of appreciation and say your plugin worked quite well in a certain project and saved me a lot of time from implementing it in MATLAB. Thanks for that :sweat_smile:

1 Like