Pipeline for Foam Bubbles?


#1

Hello,

I am analyzing the morphology of polymeric foams and trying to design an appropriate pipeline in CellProfiler to identify bubbles and the number of cell density for polymeric foaming application. I have a set of images of polymeric foams, for example, foam images were shown in kmitl.ac.th/~s7010458/foam/foam.html

We want to:

  • Identify bubbles and count these cell density
  • Measure bubbles diameter

Can you suggest a simple pipeline?

Thank you very much,

Piyapong


#2

Hello,
I am assuming that the higher intensity (brighter) circles are bubbles, but I am not sure if the dark circles are bubbles, as well. Are you hoping to identify the bright, dark , or both types of bubbles?

Please let me know, so I can help refer you to the best pipeline (and the modifications that you might need to make).
martha


#3

Hello,

I’ve plot intensity (gray scale) with imageJ, and I see each foam bubbles look like a hole. Center of every bubble are dark (slowly form bright circular bound to dark depth-center) but each bubble aren’t same dark scale, depend on bubble, some depth and some little. When we plot these images, We can observe bubbles-contour and every circular low-level are center of bubbles. I expect we can count number of bubbles with this center.

Piyapong


#4

You will certainly have to do an illumination correction to remove the varying intensity across the image. (It looks like the light source is on the bottom, left of the image. Then, I would try to identify the objects using IdentifyPrimAutomatic. To count the ‘dark bubbles’, you will need to invert the image and identify those objects.

So, your pipeline might look something like this:
LoadImages
CorrectIllumination_Calculate
CorrectIllumination_Apply
IdentifyPrimAutomatic (for higher intensity bubbles)
InvertImage
IdentifyPrimAutomatic (for dark bubbles)
Measure…


#5

Thanks,
I will try all of your recommendations and come back with results, hopefully!