Classifier using 2 channels from one image




I made a pipeline using cellprofiler and I want to analyze it using classifier.
I am new in this so I don’t know if I made the pipeline correctly.
I would like to make the analysis using image similar to the one uploaded using the two channels but when I load the image on the classifier it only shows the red channel.

Do I have to change something on the properties file or am I making a mistake in the way I save object and images to the database?



Without the pipeline and/or the properties file, there’s no way for us to know if it’s set up correctly or not, sorry!


Here are the pipeline and the properties file.
I also attached the images I am working on. The maximum images are local maximum detected using python pipeline. As I have no nuclei it acts as primary object. I am not using the green channel at all for the moment. Should I try to detect it as a tertiary object? (16.2 MB)
test_robustesse.cpproj (968.2 KB) (7.0 KB)



You don’t have to detect objects in the green channel, but you should at least 1) Load it (if it’s not part your RGB) and/or Extract it (if it is) and make some measurements in it in your Texture and ObjectIntensity modules.


Thank you for your answer. I think I did what you suggest but it didn’t solve the problem. The green channel is part of the original image so while converting to grey I extract both red and green channels separately.
I added the image in the measure texture and object intensity but it do not change my problem.
Maybe I should re-explain.
When loading the image/object on cellprofiler analyst I would like to see both channels because I need it to recognize each class
That’s what I have and I can only see the cell borders and the seeds. But I would like also to see all the green dots from the original image.

Thank you,


Hmm, I’m not sure, because I can see the green spots in that snapshot (or are those your seeds?).

If you want to have more granular control over what CPA shows you-

  • Explicitly extract BOTH OrigRed and OrigGreen in your pipeline
  • Save both individual channel images with SaveImages modules, making sure to turn on the Record the file and path information to the saved image? option
  • In ExportToDatabase, set Include information for all images, using default values? to “No” and select only the channels you care about (and what color you’d like them to be.


Yes, thank you I had indeed to save the green channel in files and provide the path to the database.