Premium
Wpt (waypoint) shift manifold difference metrics for evaluation of varying observing‐condition (observer + illuminant) metamerism and color inconstancy
Author(s) -
Derhak Maxim W.,
Luo Lin,
Berns Roy S.
Publication year - 2020
Publication title -
color research and application
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.393
H-Index - 62
eISSN - 1520-6378
pISSN - 0361-2317
DOI - 10.1002/col.22554
Subject(s) - hue , artificial intelligence , standard illuminant , computer vision , color space , mathematics , color difference , chromatic adaptation , color model , gamut , pattern recognition (psychology) , computer science , enhanced data rates for gsm evolution , image (mathematics)
Abstract Metamerism is a critical color phenomenon which can cause serious problems for products assembled by various parts. Customers generally expect all parts are color‐matched under different observing conditions. This article extends the concepts of illuminant and observer metamerism to observing‐condition metamerism, that is, objects are color‐matched under one observing condition but not under others. The color inconstancy of a single object is also expanded to be evaluated under multiple observing conditions. Moreover, four Waypoint (Wpt) Shift Manifold difference metrics are proposed to evaluate not only observing‐condition metamerism of metamers and paramers but also observing‐condition color inconstancy of single objects: The Mean Object Inconstancy Index (MOII), The Mean Object Color Difference (MMOCD), Object Metamer Index (OMI), and Object Hue Similarity Index (OHSI). Existing indices of metamerism and color inconstancy employ appearance matching using a Chromatic Adaptation Transform (CAT) and color difference formulas such as CIEDE76 or CIEDE2000. The proposed metrics utilize material matching based upon the Waypoint Material Adjustment Transform (Wpt‐MAT) and Euclidian color difference in the perceptually uniform Material Color Equivalency Space WLab. Conceptual comparisons between these approaches are discussed and evaluated. Additionally, computational evaluation results under observing conditions composed by 99 illuminants and 70 observers show that MOII provides a measure of color inconstancy for single objects, MMOCD provides a measure of metamerism between metamers and paramers with a generalized assessment of color difference between two objects, OMI provides a measure of paramers, and OHSI provides a quantitative measure of hue characteristics for different observing conditions.