Working Up

by Dwayne Phillips

The cell phone camera plus computer vision plus computer speech equals apps for the blind.

I worked in computer vision during the 1980s and early 1990s. My work was mostly with having a computer transform an aerial image into a map. I didn’t foresee the technologies that would make the digital image sensor an aid for the blind, but that is where we are today.

I recently blogged about the camera as the new I/O device. The digital “camera” (charge coupled device sensor) has become capable and inexpensive. Put a little plastic “lens” in front of it, and you have a sensor in your hand that was science fiction only a generation ago.

Now put computer vision on a computer. The computer fits in your hand and, like the camera, has science fiction power from a generation ago. The magic comes in the algorithms that computer vision researchers have developed. The software can “read” road signs and dollar bills and can tell the difference between a can of gasoline and a jug of water. Note the miracle of the Google car.

Now put some speech technology as output. This one was developed a generation ago, but we didn’t foresee it as an aid to the blind.

What do we have? Apps for the blind. The cell phone has the digital sensor, computer, computer vision algorithms, and voice output. The blind person points the cell phone about and the computer informs the person of the situation.

I must confess that I didn’t pay attention to these development and didn’t connect the pieces until about nine months ago. My granddaughter was born with a problem in her eyes. Surgery at the age of one month appears to have corrected the problem. She will not live a “normal” life, but she will see. Millions are not so fortunate as my granddaughter. I like to think that some of the work I did, some of the writing and teaching that I did, has helped advance these apps for the blind.