WebInSight: Making Web Images AccessibleBy Bigham, Jeffrey P.; Kaminsky, Ryan S.; Ladner, Richard E.; ASSETS 2006 - The Eighth International ACM SIGACCESS Conference on Computers and Accessibility,
Publication Date: 2006
Discusses the state of image accessibility on the Web and demonstrates the accuracy and utility of WebInSight, a system that automatically generates and inserts alternative text for images on web pages. WebInSight is installed as a plug-in or extension for Internet Explorer or Firefox. When a user accesses a webpage, it passes through a proxy server which arranges for images to be labeled on the user’s end. Alternative text may be drawn from an existing database of labeled images or generated on-the-fly using contextual analysis, optical character recognition (OCR), and human labeling for contextually significant images such as logos, buttons, links, and text images. Current state of image accessibility is discussed, highlighting previous studies and showing two examples of pages with inaccessible images. This discussion includes the authors’ analysis of approximately 1,000 websites for the presence of appropriate alternative text. Approximately 2,500 images with no alternative text were processed using the WebInSight system. Of those, WebInSight automatically provided alternative text for 43.2% of the images, with 94.1% accuracy. Future discussion includes a fee-based service and a tool for web authors to generate suggested text for images.
Published by: Association for Computing Machinery (Website:http://www.acm.org)
SIGACCESS (ACM Special Interest Group on Accessible Computing) (Web Site: http://www.sigaccess.org )