Web-Page Color Modification for Barrier-Free Color Vision with Genetic Algorithm

Manabu Ichikawa¹, Kiyoshi Tanaka¹, Shoji Kondo², Koji Hiroshima³, Kazuo Ichikawa³, Shoko Tanabe⁴, and Kiichiro Fukami⁴

¹Faculty of Engineering, Shinshu Unversity
4-17-1 Wakasato, Nagano-shi
Nagano, 380-8553 Japan
{manabu,ktanaka}@iplab.shinshu-u.ac.jp

²Kanto Electronics & Research, Inc.
5134 Yaho, Kunitachi-shi
Tokyo, 186-0914 Japan

³Social Insurance Chukyo Hospital
1-1-10 Sanjyo, Minami-ku, Nagoya-shi
Aichi, 457-8510 Japan

⁴Research Institute of Color Vision
12-26 Sanbonmatsu-cho, Atta-ku, Nagoya-shi
Aichi, 456-0032 Japan

Abstract. In this paper, we propose a color modification scheme for web-pages described by HTML markup language in order to realize barrier-free color vision on the internet. First, we present an abstracted image model, which describes a color image as a combination of several regions divided with color information, and define some mutual color relations between regions. Next, based on fundamental research on the anomalous color vision, we design some fitness functions to modify colors in a web-page properly and effectively. Then we solve the color modification problem, which contains complex mutual color relations, by using Genetic Algorithm. Experimental results verify that the proposed scheme can make the colors in a web-page more recognizable for anomalous vision users through not only computer simulation but also psychological experiments with them.

LNCS 2724, p. 2134 ff.

Full article in PDF

lncs@springer.de
© Springer-Verlag Berlin Heidelberg 2003