Colour discrimination scores combining colour fidelity and gamut area characteristics

Abstract

This paper investigated colour discrimination based on current available indexes, and predictors were proposed for global and targeted colour scenarios. Thirty participants conducted the Farnsworth–Munsell 100 hue test under 21 lighting conditions. The experiment results revealed that the distribution of total error score (TES) and adjusted total error score (TES_adj) showed arc shapes centred on the optimal point (100, 100) in both R_f–R_g and colour rendering index–gamut area index coordinate systems. On this basis, global colour discrimination scores, CDS1 and CDS2, based on the colour fidelity and colour gamut characteristics, were proposed. The results demonstrated that both CDS1 (r = 0.82, p < 0.001) and CDS2 (r = 0.81, p < 0.001) provided good linear correlations with TES, and CDS1 (r = 0.75, p < 0.001) and CDS2 (r = 0.77, p < 0.001) also exhibited a good linear correlation with TES_adj. Furthermore, the global colour gamut was divided into four local colour spaces (red–yellow, yellow–green, green–blue and blue–red), and the CDSs in the local gamut (CDS_local and CDS_adj,local) were constructed using the local colour properties, including R_cs,local, R_hs,local and R_f,local. The linear regression results demonstrated that CDS_local ( $\overline{r}=0.79$ ) and CDS_adj,local ( $\overline{r}=0.78$ ) can be effective colour discrimination predictors for the targeted colour scenarios.