Correction Method of Color-Matching Functions Using Genetic Algorithm and Gaussian Approximation

Abstract

Spectrocolorimeters are widely used to adjust the color of items such as LCDs, CRTs, projectors, and other video equipment. However, the color appearance of one display is sometimes mismatched with that of another display, even though the colorimetric values of the two - which are measured at the same time and under the same illumination - have been matched. It has been reported that the gap between color-matching functions and the human perception of color causes this phenomenon. The purpose of this study was to correct color-matching functions to reduce the gap between colorimetric values and human perception of color. Color-matching functions were approximated by Gaussian functions, which were controlled by ten parameters. The parameters were optimized using genetic algorithm to minimize color differences between the colorimetric values of the supervised data. Results show that the perceptual color matching was improved.