Evaluation Studies, Based On The Color Difference Of The Surface Color

Colour-difference evaluation is very important to the colour quality control of industrial products. The colour-difference formula is used to evaluate the colour difference between two colour samples, which calculated colour-difference value is based to judge whether the difference between two colours is visually acceptable. Hereby many colour-difference formulae have been proposed to simulate human eyes' perceptibility and estimation to colour differences.Most advanced colour-difference formulae, such as CMC, CIE94 and CIEDE2000, were developed upon visual assessment data sets for medium-to-small colour differences and so applied for evaluating medium-to-small colour differences, typically less than 5 CIELAB units. However, in practice, very small (below 1 CIELAB unit) to very large (above 30 CIELAB units) colour difference would be encountered in many application areas, especially in graphic industry. On the other hand, it is much complex to estimate colour differences of different magnitudes using different formulae in industrial applications, and the results calculated by one formula cannot be compared with those calculated by another formula. Hence further studies need to be carried out immediately in order to systemically analyse and evaluate the different colour-difference formulae's relativity to visual evaluation and effective prediction range, and this work is valuable in both theory and application.In this study, visual assessment experiments were specially designed and carried out for very large, medium, and very small colour-difference of object surface colour sample pairs. With the experimental data of visual assessment, some related representative colour-difference formulae were tested, which included three CIE recommended colour-difference formulae, namely CIELAB, CIE94 and CIEDE2000, the ISO standard of textile industry, CMC (l:c), and the three latest CIECAM02-based colour-difference formulae, CAM02-UCS, CAM02-LCD and CAM02-SCD. Meanwhile, the visual data were used to analyse the human eyes' acceptable colour-difference tolerance and to test the power law model. In addition, the effective prediction range of all tested colour-difference formulae were estimated, and some useful advices were suggested about how to choose appropriate colour-difference formulae to evaluate small to large colour differences in related industrial applications.Through the detailed comparison and analysis for the whole colour-difference range from very small to very large, it indicated that the three CIECAM02-based colour-difference formulae outperformed all other tested formulae. Herewith the three formulae are expected to be extensively applied to industrial applications in the future.