Low-Cost Tactile Braille Graphics

Drawings are essential in learning diagram-heavy subjects like science, math and geography. The blind experience these drawings through tactile graphics. However, these graphics are prohibitively costly to produce and prevalent in developed regions only. This collaboration between Microsoft Research India and the Mathru School for the Blind explored the potential of improving access to tactile graphics in developing regions.

The Mathru community singled out a huge gap for images in their curriculum. Intrigued, we improvised a simple experiment: using a newspaper and a Braille slate and stylus, we etched an elephant outline in dots. At first, a group of blind students was puzzled. When they understood that these dots represented the figure of an elephant and were not to be read as text, a light-bulb moment followed. They spiritedly recalled different parts of an elephant’s body and identified them in the dotted outline.

An enthusiastic response to an early experiment got us interested. How valuable could these graphics be?

Fueled by the students’ enthusiasm, we got to work at the lab. The school’s Braille printer could only print text. It limited where dots could be embossed and lacked the granularity found in most image embossers. Perhaps we could trick the printer into creating tactile graphics? Some technical puzzles later, we created a software tool that enabled this.

1. An original image showing a textbook map of India

2. Resizing the image to an 84x84 pixel frame lets us map it directly to dots on a Braille sheet

3. An iterative thresholding technique is used to bring out the dominant features of the image

4. Each 3x2 block of pixels is mapped to a Braille character to create a text file with “Braille ASCII art"

5. The text file is input to the school’s Braille printer and out comes the tactile image

At one point, we had a choice between optimizing the tool and exploring the value of these images. We chose the latter. It was a worthwhile decision cause we were in the field actively learning what worked and what didn’t. It meant shortcutting some of the development: when the algorithm didn’t generate an “optimal” image, we manually tweaked the output to get the image we wanted.

A variety of graphics, including shapes and diagrams with legends, were tested through four phases of open-ended interviews and exploratory teaching activities with students and teachers. The students were highly enthusiastic, quickly able to understand, and, fascinatingly, even learn how to handwrite the alphabet using them! Mathru printed them out into books and continues to use them to this day.

In the “handwriting" experiment, we showed a blind student how to identify the shape of letters and use a pen

Here, she feels the image of the letter ‘A' to sense its overall shape

Using one hand to hold the pen and the other to track where the horizontal line should be drawn, she successfully writes the letter

She went on to write the entire English alphabet by hand...

...and teach her classmate how to do the same

Motivated by the potential to sign checks at banks, this teacher used the images to teach herself how to handwrite in a week!

The project showed me how technology could make a positive impact in someone’s life. By immersing ourselves in the lives of our users, we gained a priceless appreciation of their aspirations, attitudes and ordeals. I was first exposed to prototyping in this project: building something to test an idea instead of building it for production. This was also the first time I published and presented a paper at a major conference.

Presenting our work at the ACM DEV 2010 conference at the University of London, UK.

SCOPE

Figure out an inexpensive way to generate tactile Braille images and measure the effectiveness of these images

PERSONAL IMPACT

Prototyping, user research

ORGANIZATION

Mathru School for the Blind, Microsoft Research India

team

Saurabh Sanghvi, Kentaro Toyama

TEAM DISCIPLINE

Computer Science, Electrical & Computer Engineering

location

Bangalore, India

DURATION

May 2008-August 2008

Deliverables

Software tool, library of 50+ tactile Braille images

TOOLS USED

publication

ACM DEV ’10 Proceedings of the First ACM Symposium on Computing for Development, London, UK, December 2010

TYPE

Information and Communication Technologies and Development for developing regions (ICTD)