Eye Tracking Research Exploring Assistive Possibilities
Many of our posts revolve around the ever expanding industry of usabilty testing with eye tracking technology. So it’s a pleasure to see more use of eye tracking for humanitarian purposes, such as aiding in communication or transportation for those with ability limitations.
In a recent paper published at the University of Southern California – Los Angeles (USC), researchers are using eye tracking to aid with communication, facilitating disabled individuals in their everyday lives. They’ve implemented gaze tracking for a selection of options displayed on a computer screen. Their goal is do assist people with limited motor skills by providing them with the ability to communicate the desire for various basic necessities such as food and drink.
Using a Texas Instruments C6416 fixed-point digital signal processor, they’ve adapted a version of a specialized algorithm for implementation. Oftentimes, eye tracking is done with the help of infrared illumination, but in this case, researchers have chosen to use visible spectrum illumination, which enables them to concentrate on tracking the limbus contour. When using visible spectrum illumination, it’s possible to track the iris-sclera boundary, also known as the limbus. At the limbus boundary, the luminescence gradient is high, allowing detection by simple gradient algorithms.
The researchers opted not to supplement the study with head tracking, as head tracking systems tend to be more expensive than stationary cameras. The user sits still and focuses on a table-mounted screen from which they can choose one of four choices by staring at any corner of the computer screen.
The researchers from the Ming Hsieh Department of Electrical Engineering use a variety of back-end features such as Region of Interest(ROI) Locatlization, essentially narrowing the camera’s search space or field of view to focus more specifically on a particular area; Ray Tracing, a system based on the Starburst algorithm where rays are traced radially and outward along predefined paths from the center point; Circle Fitting, using an algorithm to compute the radius and center point of the circle to best fit the data; and Center-Point Identification/Quadrant Mapping, getting an accurate reference point from which to begin the measurements.
Essentially an experiment, the researchers found there needs to be a few changes before the system can be considered for market – lower price, less “intrusive”, and they need to solve the problems of limited tracking resolution and movement compensation.
Eye Catchers: A System for Eye Gaze Detection Using the TI C6416 DSP
Related articles:
- Eye Controlled Video Games? Better Late Than Never
- Eye Tracking: Viewing the World Through Someone Else’s Eyes
- Eye Tracking Research Measures Your Problem Solving Skills
- Eye Tracking Research Measures Effectiveness of Super Bowl Ads
- Eye Tracking Exploring the Differences Between Baidu and Google
- Eye Tracking Technology as Memory Aid
- The Latest in Eye Tracking Web Usability Research pt1
- Eye Tracking in New Interfaces
- Eye-Com Eye Tracking Drive Simulator Studies Driver Inattention
- Eye Tracking History: An Early Eye Tracking Apparatus