Optimizing settings to Identify Objects




I am using the IdentifyPrimAutomatic Module to count the number of beads in the Images. The beads are clumped together and there are water bubble in the images. Even after modifying the settings the module is still not able to totally count the total number of beads. I have posted the images that I have obatined after using the module.


here are my settings for the module
diameter — 18,37
Thresholding – MoG Global
Threshold correction factor – 1
Lower & Upper bound on threshold – 0,1
Fraction of image covered by objects – 0.2
methods to distinguish clumped objects – Shape
Methods to Draw dividing lines between clumped objects – Intensity
Size of smoothing filter – 2.5
Supress local maxima – 1
speed up local maxima within the distance – 1

Thanks for the generous help



Thanks for posting a general outline of your pipeline. There are a few suggestions that I have for you:
~Method for diving lines between clumped objects --try Distance
~size of smoothing filter increase to 37. This way, the smoothing size is closer to the size of the bead diameter.

Remember, it is hard for CellProfiler to detect beads if you cannot detect the edges, either. With that in mind, it is hard for the software to detect the beads around the the water bubble.

Hope this helps!