Detecting red yeast colonies


I’ve been trying to count total, white and red colonies, but the red colonies are not recognized using the identify primary object function. When I look at the images following convert to gray, the red colonies are not visible in the red channel and are appearing black in the green channel. Is there some way to correct this? I have attached an image of a plate as an example.

Thanks for your help.

Hi Tricia,

Have you taken a look at our yeast colony counting pipeline on our examples page? It seems similar to what you have, so you may want to try it as a starting point. However, the main caveat is that since your images are not exactly the same as the example, the pipeline will probably not work without some adjustments on your end.


Hi Mark,

Thanks for the reply. I started with the pipeline and made adjustments to allow colonies to be picked up. However, the reds are not being identified because they do not appear in the red channel (I suspect that the difference between our images and yours is the media. We are using rich media,which has a yellow tint, and your sample plate is minimal media, which is colorless). In any case, I’ve tried to change the identify primary objects part to look at the other images (blue and green channel), but it never picks up the reds (they are black in those channels). Do you have any suggestions for how to further modify the pipeline?

Thanks for your help.

I noticed that if you invert the green channel with ImageMath, your red colonies stand out as bright objects that you might be able to segment with IdentifyPrimaryObjects. Multiplying the inverted green channel by the inverted blue channel may make the effect stronger.