Figure 3.

Definition of our colour constancy index. The CIE coordinates of the Munsell paper under the experimental illumination (cross; ie, an absence of colour constancy), the correct colour match under the lamp illuminating the Munsell book (open triangle; ie, perfect colour constancy), and the corresponding observer colour match (open square) are plotted in the CIE (x, y) chromaticity diagram. The vector ‘a’ represent the Euclidean distance for the CIE y coordinate between the observer's colour match and the absence of colour constancy (‘match of reflected light’). This vector represents the shift in the observer's colour perception caused by the experimental illumination. Vecor ‘b’ represents the shift in chromaticity (for the CIE y coordinate) of the Munsell paper caused by the illumination (ie, the amount of illumination change that the subject has to discount). Vectors ‘c’ and ‘d’ show the same principle for the CIE × coordinate. Our colour constancy (CI) index is defined by:

$${\displaystyle \mathrm{CI}=\frac{\left|\mathrm{d}\right|\ast \left|\mathrm{c}\right|-\left|\mathrm{b}\right|\ast \left|\mathrm{a}\right|}{{\left|\mathrm{d}\right|}^2+{\left|\mathrm{b}\right|}^2}}$$