Applied perspective on color difference measurement

Abstract

Recent advances in electronic display and hard copy technology have made the use of full-color presentations feasible for a variety of applications. Color can enhance pictorial realism, increase human operator visual throughput, and facilitate performance in complex task environments. Realization of these benefits requires the specification of discriminable color sets and reliable tolerances on system visual parameters. The measurement and prediction of color differences are fundamental to these processes. This paper examines the use of existing measures of small color differences, such as CIELUV and CIELAB, for color specification problems for electronic visual displays. While small color difference formulas appear intuitively appropriate for specifying chromaticity tolerances for visual displays, their a priori applicability to large color difference problems such as color set selection would seem questionable. Nevertheless, small color difference formulas have been used for display color selection with considerable success. This success is manifest in consistent relationships between color difference estimates and a variety of visual performance measures, including color discrimination, color identification, reaction time, and visual search. Practical limitations of existing approaches to color difference measurement are discussed within the context of today’s complex color imaging devices and varied application environments.