UMBC ebiquity

Recoloring Web Pages For Color Vision Deficiency Users

Authors: Vikas Bansal

Date: July 11, 2014

Abstract: Color vision begins with the activation of cone cells. When one of the cone cells dysfunction, color vision deciency (CVD) ensues. Due to CVD, users become unable to differentiate as many colors as a normal person can. Lack of this ability can lead to less rich web experience, incomprehension of basic information and even frustration. Solutions such as carefully choosing colors during the design or recoloring of web pages for CVD users exist. Nevertheless, extant solutions are limited in that: 1) they overlook the varying strength in terms of the relationships between pairs of colors from the same websites, and 2) they assume that the original web designers have followed the recommended guidelines with respect to the contrast ratio of related colors. To address the above limitations, we propose an improvement to a state-of-the-art recoloring tool in terms of the time complexity. More importantly, we present our tool, FBRecolor, which leverages the foreground-background relationship between colors in a web page to further improve the recoloring performance while preserving naturalness, pair-differentiability and subjectivity. Furthermore, we introduce a new component into FBRecolor that takes into account of color contrast in selecting replacement colors according to W3C guidelines. The experiment results show that FBRecolor does better in preserving pair-differentiability and produces lower total cost solutions than SPRWeb. In addition, the results demonstrate that the proposed extension to FBRecolor improves the color contrast ratio of selected replacement colors for problematic websites.

Type: MastersThesis

Organization: Computer Science and Electrical Engineering

Publisher: University of Maryland, Baltimore County

Google Scholar: search

Number of downloads: 471


Available for download as

size: 2040898 bytes