| |
|
|
Eye Tracking
The pattern of a person’s eye movements over an image can reveal whether the person has seen the image before. This is especially true for photos of faces, but has also been shown for pictures of scenes. Movement of the eyes can be monitored by illuminating them with an eye-safe infrared light which creates a reflection that can be detected by a special camera. Software triangulates the reflection to calculate where the eyes are focused on an image, and tracks the focal point as the eyes move. Specially designed algorithms use the eye movement information to determine whether the image is familiar or unfamiliar. Research has found accuracy above 85% for recognition of familiar faces, with lower accuracy for other types of stimuli. While the technology is non-contact, the rigorous testing protocol limits its usefulness in a covert capacity. This technology is approaching maturity, and may be available to law enforcement agencies in the near term.
|
|
|
|
|
 |
