FISH - Fluorescence in situ hybridization - counting the cells

Hi guys,
I am new in this program, but I think that it can helps me a lot with my task. My experiment is FISH - fluorescence in situ hybridization. Basically, I have 1 OIB picture from Olympus confocal microscope, which contain 3 channels. First channel is for DAPI (blue) - all bacteria cells should be stained. Second is Eubacteria (EUB, green) - some of the bacteria are stained. And third channel is for specific probe (red) - only a few cells are stained.

So all I need is count the total number of DAPI (all the cells), then the total number of EUB (cells which are stained by DAPI and EUB) and finally total number of cells which are stained by specific probe (these cells must be stained by DAPI, EUB and specific probe).

Problem is that the pictures are not so clear and Cellprofiler does not support OIB.

I was working on one pipeline, but it is not working correctly I think.
At the beginning I found this topic: Triple positives and nuclei texture
which is kind of similar to my problem.
So I used this pipeline and transform it. I used the ImageJ to split the channels into 3 TIFF pictures so there should be no problem for Cellprofiler (if there is any other way, please tell me).

I have lots of samples so it would be great to find a way how to count the cells easily. I am sure that the Cellprofiler can make it! :wink:

I uploaded the upgraided pipeline and the file with OIB pictures and some pictures, which I split beside the channels.

////https://onedrive.live.com/redir?resid=672F5683BEA122CB!22529&authkey=!APuNucDO77QK7GY&ithint=folder%2C//// - I have to put inside the / / because as a new user I am not alloweded to use more than 2 links.

FISH_V1.cppipe (39.5 KB)