I used ImageJ software to quantify ultrasound color-coded images, I think you might be able to use my protocol, which could be found at two different sources:

https://www.ncbi.nlm.nih.gov/pubmed/?term=najla+alsiri

http://eprints.uwe.ac.uk/30112/18/Thesis_Najla_4RegistoryV2.pdf

Thank You Najla Alsiri

If they're not fluorescent you can't really quantify them, as colour is non linear, subjective and is dependent on the specific light source used. There's an imageJ plugin that allows you to detect relative proportions of each colour from within 1 image that is sometimes misused by people who haven't read the plugin documentation. If you want to quantify fluorescence is your best bet.

Thank You Howard Vindin