refract.me

For an NSF-sponsored REU at UC Santa Cruz, I did an evaluation examination as to whether comics could be made accessible to people who are visually impaired.

Why?

• Access to information, particularly visual information, is still very limited.
• The only text-image conversion available is for semi-automated conversion of abstract diagrams (histograms, flow charts, etc.).
• Comics have mingled text and images.
• Images are fantastical and unrealistic in comics (allowing a level of abstraction and reduced complexity).
• Comics limit text to very small areas.

How?

I created an image conversion pipeline (which I controlled manually) that divided the work into creating a tactile image and producing an audio overlay.

Image conversion

It’s a fairly straight-forward process, and could easily be automated.

Audio Overlay

The audio was connected to the embossed images using a Talking Tactile Tablet (TTT). These areas are fairly large, and can have multiple levels of audio.

This was more time-consuming and the TTT is a fairly expensive piece of hardware.

More Information

• My full final report.
• SURF-IT REU site.
• My SURF-IT page.
• Roberto Manduchi, my advisor.