Yuta Asano


Colors are typically described by three values such as RGB, XYZ, and HSV. This is rooted to the fact that humans possess three types of photoreceptors under photopic conditions, and human color vision can be characterized by a set of three color matching functions (CMFs). CMFs integrate spectra to produce three colorimetric values that are related to visual responses. In reality, large variations in CMFs exist among color-normal populations. Thus, a pair of two spectrally different stimuli might be a match for one person but a mismatch for another person, also known as observer metamerism.

Observer metamerism is a serious issue in color-critical applications such as soft proofing in graphic arts and color grading in digital cinema, where colors are compared on different displays. Due to observer metamerism, calibrated displays might not appear correctly, and one person might disagree with color adjustments made by another person. The recent advent of wide color gamut display technologies (e.g., LEDs, OLEDs, lasers, and Quantum Dots) has made observer metamerism even more serious due to their spectrally narrow primaries. The variations among normal color vision and observer metamerism have been overlooked for many years. The current typical color imaging workflow uses a single standard observer assuming all the color-normal people possess the same CMFs. This dissertation provides a possible solution for observer metamerism in color-critical applications by personalized color imaging introducing individual colorimetric observers.

In this dissertation, at first, color matching data were collected to derive and validate CMFs for individual colorimetric observers. The data from 151 color-normal observers were obtained at four different locations. Second, two types of individual colorimetric observer functions were derived and validated. One is an individual colorimetric observer model, an extension of the CIE 2006 physiological observer incorporating eight physiological parameters to model individuals in addition to age and field size inputs. The other is a set of categorical observer functions providing a more convenient approach towards the personalized color imaging. Third, two workflows were proposed to characterize human color vision: one using a nomaloscope and the other using proposed spectral pseudoisochromatic images. Finally, the personalized color imaging was evaluated in a color image matching study on an LCD monitor and a laser projector and in a perceived color difference study on a SHARP Quattron display. The personalized color imaging was implemented using a newly introduced ICC profile, iccMAX.

Library of Congress Subject Headings

Color vision; Colorimetry; Visual perception

Publication Date


Document Type


Student Type


Degree Name

Color Science (Ph.D.)


Mark D. Fairchild

Advisor/Committee Member

George Thurston

Advisor/Committee Member

Roy S. Berns


Physical copy available from RIT's Wallace Library at QP483 .A73 2015


RIT – Main Campus

Plan Codes