CIE defined color matching functions (CMF) for XYZ and thus RGB colors. But of course, each camera has its own color sensitivities for visible spectrum. Is there any way to estimate this function (or a series of values by wavelengths) for a given camera?
closed as off topic by Qmechanic♦, Manishearth♦, Ebenezer Sklivvze♦ Dec 29 '12 at 15:16
Questions on Physics Stack Exchange are expected to relate to physics within the scope defined by the community. Consider editing the question or leaving comments for improvement if you believe the question can be reworded to fit within the scope. Read more about reopening questions here.If this question can be reworded to fit the rules in the help center, please edit the question.
Take image of 18% Kodak grey card - or just a piece of normal white office paper. Measure relative signal in red/green/blue channels.
In reality you would need a monochromator and lightsource to measure the full spectral resposnse but since there is nothing you can do to correct the shape of the RGB filters or their overlap there isn't much point.
If you know the spectral response of the camera in detail AND the spectra of the light source + reflectivity object you could calculate a correction - but if you knew all that you wouldn't need a colour camera.