Braigo launches web app to help blind people parse text on images

Picture of Shubham Banerjee, founder of Braigo Labs, working with a soldering iron. The image is overlaid with the text "Your screen reader can't read this text".

The internet is all about sharing information, but a lot of the information shared assumes that its users are able to rely on their senses to consume said information. That isn’t always the case, and for the more than 285 million people around the world who are visually impaired, browsing the ‘net can be a less than fruitful experience. Braigo Labs is today launching a beta of its service to help blind and visually impaired people make sense of text on images. The beta is the first part of a bigger platform, taking a broad-spectrum approach to helping those unable to see.

Braigo Platform is a free web application aimed at anyone who needs accessibility solutions for the visually impaired. The platform supports more than 50 languages, and can be used to extract text from images from a variety of sources, whether from the web or from a phone.

Screen shot of a Google Image Search page showing photos with image macros

Image macros or memes are particularly problematic, the Braigo team tells me.

“This is our step towards the right direction,” says Shubham Banerjee, founder of Braigo Labs. “We are looking at Braigo as a whole product that creates, supports and expands an ecosystem for the visually impaired. Not simply as a standalone software or hardware tool.”

Screen readers have been available for a while, but are usually limited to text displayed as text (much like the words you are reading now). They are less useful for reading text overlaid on images, or, say, screenshots of text. New APIs and services are popping up all the time to help computers understand images, which will undoubtedly be helpful to visually impaired people, as well.

Some of this app-enabled tech has been around for a while, and Braigo is launching itself into a market that has quite a few companies in it already, including the KNFB reader app that can help read pretty much any text out loud, LookTel’s Money Reader that helps identify currency, the open-source TapTapSee app, which can identify objects and, of course, the Be My Eyes initiative, which crowdsources able-visioned people to help blind people see.

Browsing while blind

If you’ve never tried surfing the web using a screen reader, you’ve probably not experienced that trying to navigate the web with serious visual impairments can be a hit and miss experience.

Ah, who are we kidding — it’s mostly a complete farce. Don’t get me wrong, there is no shortage of guidelines for how to make the web more accessible to blind and visually impaired users, but a lot of web developers seem to simply not care. Even some of the most frequently used websites out there fail at the very basic task of explaining what an image is. Don’t believe me? Try this example…

Screen shot of the IRS website, with its source code next to it

Death and taxes, and all that: The IRS is arguably one of the most important web sites in the U.S. The most important thing you’ll do is to file your tax return, which is why the IRS website has that friendly smiling man front and center. Unfortunately, the IRS also decided to put the text (“File Your Tax Return”) as part of the image. That wouldn’t have been a problem, had they decided to use an alt attribute to describe what the image is. A simple alt=”Button to file your tax return” would have done the trick, but instead, Uncle Sam decided to use a title tag. Lovely, but no good to people relying on screen readers.

Continuing the work

If Shubham Banerjee’s name rings a bell, that’s because he was the 13-year-old entrepreneur who raised investment from Intel Capital back in 2014 to help further develop his Lego Braille printer. The video below shows how that project came together a couple of years ago.