Supplementary dataset for color difference evaluation

Abstract

The Technical Committee 1-55 of the CIE on ''Uniform color space for industrial color difference evaluation'' requested in 2007 reliable experimental datasets (that is, color pairs assessed by a considerable number of observers with nondefective color vision, under well controlled experimental conditions and using proper experimental methods). The aim was to use these datasets to develop/test new color spaces with Euclidean color-difference formulas more robust than the CIEDE2000 color-difference formula. The objective of this paper is to propose a new dataset more reliable than previous datasets and to evaluate from this dataset more accurate just noticeable differences. In our study we have considered not only surface specimens (reflective objects), but also mainly color pairs assessed using visual displays (e.g., self-luminous LCD). Our experimental results have been obtained under illuminating/viewing conditions close to the ''reference conditions'' suggested for the CIEDE2000 color-difference formula (CIE Publication 142-2001) and by the ISO standard 3668:20033668: (ISO standard 3668:2003 Paints and varnishes -Visual comparison of the colour of paints).

Analyzing small suprathreshold differences of LCD-generated colors

“…Previous investigations of small suprathreshold color differences viewed on an LCD focused mainly on their absolute deviations from surface-color differences and from the predictions of color-difference formulas [12]. Other experiments conducted with LCDs and surface colors analyzed justnoticeable color differences and the performance of color-difference formulas in predicting such differences [13,14]. Parametric effects influencing the perception of small color differences, such as background color, psychophysical method, or separation of samples, were investigated using CRT or surface-color stimuli [15,16].…”