Segmentation on color and calibration

Can Cell Profiler segment based on color ? If so, would you please indicate the block to use ?

Can a set of images be calibrated so that area are in eg. sq microns vs pixels ?

Regards,
Paul

Hi,

Strictly speaking, no, the segmentation in CP is done on grayscale images. However, if you mean true color images, e.g. histology images, then you can use UnmixColors and segment the individual grayscale channels you create. Otherwise use ColorToGray to separate, say, a 3-color RGB image into the 3 original channels.

From FAQ:
viewtopic.php?f=14&t=806#p4564

Cheers,
David

Thanks for the prompt response. I’ve been using colors to gray and will try Unmix.

btw, comparable results between IJ and CP were obtained in segmentation on a set of 110 slides of a degradable implant. The only difference was that analysis in one case took 3 hrs, in the other 3 min.

I’ll check the FAQ

Paul

[quote=“pbuscemi”]btw, comparable results between IJ and CP were obtained in segmentation on a set of 110 slides of a degradable implant. The only difference was that analysis in one case took 3 hrs, in the other 3 min.
[/quote]

I hesitate to ask, but which was which?