Assessment of Color‐Measuring Instruments

Abstract

The U.S. Army has undertaken a program to develop an instrumental method for assessing the acceptability of textiles for color difference from a standard. This article reports the results of the first phase of the program, an assessment of three commercial color‐measuring instruments (Diano Match‐Scan, Hunter D‐54P‐5, Macbeth MS‐2000) for objective textile acceptability judgment. It is concluded that the three instruments are essentially equivalent in the precision and accuracy of the measurement of color and color difference using a wide variety of samples, at least over periods of up to seven weeks. All measures of repeatability lead to the conclusion that the uncertainties involved are well below the just‐perceptible color difference. Levels of absolute accuracy achieved depend greatly on the details of operation, data treatment, and calibration, but are considered satisfactory for each of the instruments tested. With respect to certain other parameters, the performance of the instruments is less satisfactory. Rejection of the specular component differs significantly among the three, as does the selection of weights for tristimuus calculations. One instrument, as tested, exhibited significant sensitivity to weave orientation in textile samples. Finally, we find that the distributions of tristimulus values obtained with each of the instruments show large deviations from normality, severely limiting the significance of conventional statistical treatment of such data.

Evaluation of methods for verifying the performance of color‐measuring instruments. Part I: Repeatability

“…One important conclusion is that traditional color difference calculations (ie: DE* ab and all its univariate derivatives) are not distributed as normal random variables. 7 Hence statistics such as variance and standard deviation, often applied to color difference data, are not strictly appropriate for these data.…”

Four color-measuring devices compared with a spectrophotometric reference system

“…For optimum results, these devices should be sufficiently accurate and should achieve high repeatability with low measurement uncertainty. Numerous studies have analyzed the repeatability [14][15][16][17][18][19][20][21] and accuracy [22][23][24][25][26][27] of commercial color-measuring devices. This in vitro study compared the L*C*h8 color coordinates of four dental color-measuring devices with those of a spectrophotometric reference system.…”

Visual gamma correction for LCD displays

“…To calculate both variability measures, each digital rgb value is transformed into device-independent CIE XYZ tristimulus values by using the ground truth measurements obtained with the spectroradiometer. Then the CIEDE2000 colour difference formula [12] was used to calculate the mean colour difference to the mean value (MCDM) [13], for each colour channel for inter-observer variability. For intraobserver variability, the CIEDE2000 colour difference between each observer's judgments in session 1 and in session 2 was calculated and averaged for each colour channel.…”