Illumination correction required but facing a Memory issue


I am trying to account for the irregular Illumination in each plane(channel) of a RGB image using the ‘CorrectIlluminationCalculate’ module.
My images are nearly 7Mb and are of dimensions 2848 * 4288.
This obviously lead to memory errors.
I tried using a few tips from this post: CellProfiler 2.0 FAQ

I have attached my pipeline as a JPEG image.
I have split the channels, applied the illumination correction individually for each channels and then converting to the truecolor image and saving them. I use conserve memory module to get rid of the unwanted intermediate images.

I still am facing the memory error.
How do I get over this problem? Is there a way where I could increase the amount of memory cell profiler uses?

I am also interested to know about the algorithm used in CorrectIlluminationCalculate module.
The manual does not have much information about the algorithm.
I tried reading the code but a short algorithm/suitable references would help me to understand better.
Where can I get to read more about it?

Appreciate any help.


I notice that you are using ConserveMemory after all the illumination correction is done, so you still have all the images stored in memory up to that point. Have you considered using a ConserveMemory module immediately after each CorrectIlluminationApply to remove the unneeded images right then?