Dear Rasool,

In effect the four models quoted are all either stimulus counts or colour order systems, not perception models. In order to model perception it is necessary, firstly to quantify the human visual response to light stimuli; and secondly to order the colour identities thus identified into an appropriate ‘colour space’ model.

The CIE Standard Observer XYZ Tristimulus Model which is widely used industrially for colour match prediction meets the first requirement. In each case the input to the calculation is a spectrally specified stimulus power distribution for the light in question and the output is a set XYZ colour identity co-ordinates. The second stage models such as CIELAB are firstly used to distribute the XYZ specified colours into a more easy to visualize colour space; and secondly to establish a measure of colour difference between a given pair of samples.

Thus, R G and B (Red, Green and Blue) may be additive colour mixture of primary stimuli or screen drive values which in addition sum to a white light. Cyan Magenta and Yellow (CMY) are the subtractive primaries, respectively Cyan is a minus red colorant, Magenta is minus Green and Yellow is minus blue and sum together to remove all light i.e. black. CMYK is an enhanced black coloration common in printed images where a naturally black pigment is added to a CMY subtractive combination. Finally, HSI stands for Hue Saturation and Intensity and are a set of numerically derived colour co-ordinates that are intended to specify a more visually intuitive colour space than CIE XYZ or CIELAB.