According to the December 2010 “VA Research Currents”, the Veterans Administration is exploring new technologies to help people with vision loss. Today, the blind are able to use “talking” handheld GPS devices sometimes with a guide dog to help them navigate.
However, there are limitations in using GPS devices. For example, directions for pedestrians can be off by 50 or even 100 feet, clouds or tall buildings can block signals, indoors GPS devices may not work at all, and even under ideal conditions, a consumer GPS device is usually accurate to only about 10 feet. To overcome these problems, a VA-funded group of researchers are designing a computer vision system to bridge these limitations.
“We envision combining out system with technologies such as GPS,” explained Cha-Min Tang MD, PhD, at the Baltimore VA Medical Center and the University of Maryland. He is being helped by David Ross Med, MSEE a VA rehabilitation engineer at the VA’s Atlanta Vision Loss Center.
To study the problems, a person with vision loss will wear stereo headphones with a small webcam and microphone attached to their clothing. The devices are wired to a small laptop carried in a backpack and the researchers are looking to the future when smartphones may be able to handle the computing. When the user wants to find a specific location, the computer compares the webcam’s view with still images of the area around the target that has been preloaded into the computer. Beeps and other audio signals in stereo indicate how to proceed along with computer-generated speech to provide additional feedback.
For indoor navigation, still images will be needed for every 15 to 20 feet along each path the user might follow. The idea is that a sighted volunteer would snap the images ahead of time and upload them into the user’s computer. Eventually, large public sites could eventually offer downloadable libraries containing images of high traffic areas.
For outdoors, a different approach is needed. One idea is to rely on the GPS to let the user know roughly where he is and then call up a small set of relevant images for that location. At that point, computer vision would take over and give more detailed guidance.
Another option is to add an Inertial Navigation Unit (INU) about the size of a flash drive. INUs are used to aid navigation on airplanes and submarines. The idea is to use a Google map and the GPS to let you know more or less where you are and then the INU and the camera will tell you how you are moving and pinpoint the location.
Researchers at the Atlanta VA are evaluating an alternative approach. Their system relies on a smartphone to stream video frames to a central server. The server analyzes the images and sends back navigation data. The plus is that users don’t have to carry their own computer. The minus is that connection speeds can affect how fast the system works.
The different approaches each have advantages but the researchers think that the ultimate solution is to build a combined system using complementary technologies that will be reliable under a wide range of conditions.