On Jun 24, 11:16 am, "M. Funkibut" <[EMAIL PROTECTED]> wrote:
> I assume there is a 'scanner' [some sort of filtered photoelectric
> eye] that outputs an RGB triple to a software routine that then
> produces a paint 'formula' that hardwarestoreguy can mix up.
>
> But how do you get from RGB to the paint formula?  Can you go from a
> paint formula to RGB? Does anybody know where I can find a reference
> book/article on this? I've looked and come up empty.

My understanding is that most types of house paint consist of a white
(titanium dioxide) base, and some number of pigments. This results in
a subtractive color model, like CMYK, though I believe most house
paint manufacturers use more than just 4 pigments (which could mean
that there are multiple ways to arrive at approximately the same RGB
value). It should be possible to just generalize a standard RGB to
CMYK conversion algorithm to work with an arbitrary set of pigments,
assuming the pigments can cover the entire RGB gamut.

Another thought: perhaps they don't go through RGB at all. If you have
a black-and-white camera (or even just a light meter, since they
really just need a spot measurement) you could choose a set of filters
(or colored lights) to use a different additive color model. You'd
still have to convert from an additive (light) to subtractive
(pigment) model, but there are probably some additive models that
would produce better results. For example, if the paint has 8
different pigments it'd probably be better to use 8 different filters/
lights. My guess is that colors which are the inverse of each pigment
color would work best. (just like CMY is the inverse of RGB)
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Algorithm Geeks" group.
To post to this group, send email to algogeeks@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/algogeeks
-~----------~----~----~----~------~----~------~--~---

Reply via email to