Unwanted parts in Threshold adjustment

Hi everyone
I have a question and I am looking forward to receiving your suggestions.
After applying the Threshold adjustment, some unwanted parts appear that I want to remove before the image analysis. How can I remove these parts after the threshold adjustment?
Thanks all

Hi
@mirzaeeramiiin
For the image analysis you can adjust according to the circularity and the size:

Read this:
https://imagej.nih.gov/ij/docs/guide/146-30.html#toc-Section-30
Greetings

Yes, but the problem is that those unwanted parts are in the range of the wanted circularity and particle size that must be considered

@mirzaeeramiiin
Can you drop us two pictures:

  • an annotated 'indicating what is good and what is not good)
  • the other not annotated
    Maybe we can help.

For example, the areas that are represented by red unfilled circles are those unwanted particles.
Threshold.tif (55.8 KB) Bandpass Applied.tif (574.0 KB)

Hi @mirzaeeramiiin,
Those unwanted areas seem to be a shadow from the curved area next to it.
One way I could suggest is using Ilastik. You could define different classes by pixel based classification, and then take those classified areas to the CellProfiler and make your measurements.
Hope this helps!

Hi
@mirzaeeramiiin

If this macro-code fairly cuts black holes, it does not cut clear circular things. I’ll see it tomorrow.
PS: Is it a photo of a biological structure? Something else? I’m curious to know.

run("Duplicate...", " ");
run("Duplicate...", " ");
run("Multiply...", "value=2");
run("Enhance Contrast...", "saturated=10");
setAutoThreshold("MaxEntropy");
//run("Threshold...");
//setThreshold(0, 124);
setOption("BlackBackground", true);
run("Convert to Mask");
run("Set Scale...", "distance=74.0002 known=1 unit=µm");
run("Set Measurements...", "area redirect=None decimal=2");
run("Analyze Particles...", "size=100-Infinity pixel circularity=0.40-1.00 show=[Overlay Masks] display add");

Hi @mirzaeeramiiin,
This could be addressed in couple of ways that I could think of,

  1. With Cellprofiler, (i) you could filter it using more than one measurement i.e, Area & enccentricity
    (ii) manually filter (iii)might have to get rid of these shadow before thersholding using “DIC” option in “enhanceSuppress” Module. (iv) using illastik as mentioned in the thread earlier
  2. Using ImageJ/Fiji, during thersholding & measurments later.
    Regards,
    Lakshmi
    Fujifilm Wako Automation (Consultant)
    www.wakoautomation.com
    For CellProfiler training or optimised pipeline write to,
    lakshmi.balasubramanian.contractor@fujifilm.com

Read more on our site.
Yokogawa CV8000 - The Ultimate in Confocal HCS
https://www.wakoautomation.com/products/yokogawa-high-content-imaging