Universal Design for Web Applications

Images

In HTML, the alt attribute on the img element may be considered metadata for some; for others, it is the content. For images used as links or buttons, lack of metadata creates an insurmountable barrier to people using screen readers.

We (and here we mean the big “We”—accessibility advocates around the globe) have been talking about accompanying images with brief textual alternatives (“alt text”) since September 1993, when Frans van Hoesel proposed the alt attribute, seven months after the first discussions of the img element (img: http://1997.webhistory.org/www.lists/www-talk.1993q1/0182.html and alt: http://ksi.cpsc.ucalgary.ca/archives/WWW-TALK/www-talk-1993q3.messages/983.html). Fifteen years later, images appear on almost every website but it’s estimated that only 7% provide adequate alt text for images.^[13] That’s disappointing.

So, folks are taking the issue into their own hands. Services such as WebVisum (http://www.webvisum.com/) are using optical character recognition (OCR) to ferret out the text within pictures of text (raster-based images). The technology is so good that it can solve CAPTCHAs—those twisty, hard-to-read words that are supposed to test whether you are human or a spambot trying to enter a site. Unfortunately, the primary test has become how well you can read twisty, hard-to-read words, creating a barrier for people who can’t see or read well—until now.

This isn’t to say that you are off the hook for providing alt text. People using ...