Solving crimes from retinal reflections

It is possible to identify the faces of people standing outside the frame from the reflections in the eyes of the person being photographed, a discovery that could one day help solve criminal cases.

It is possible to identify other people's faces from the reflection in the eyes of the person being photographed, a discovery that could one day help solve criminal cases.

It is possible to identify other people's faces from the reflection in the eyes of the person being photographed, a discovery that could one day help solve criminal cases.

The eyes are not only a mirror of the soul, but also a mirror of the world around us. Recently, a team of British researchers discovered that high-resolution digital images are so “sharp” that they can zoom in on the eyes in photographs and capture images of people standing outside the frame reflected on the retinas of the people in the photos.

Combine this with the fact that humans are very good at recognizing faces, even from poor quality photos, and you have a valuable source of information for solving crimes.

Previously, this assumption was considered valid if you were the one holding the camera, so you would not be in the photo.

“The pupil of the eye is like a black mirror,” says Dr Rob Jenkins, of the Department of Psychology at the University of York. “To increase the image, you have to shrink it and adjust the contrast. The image of a face that is detected by the reflection in a person’s eye is about 30,000 times smaller than the person’s own face.”

In the study, he and his colleague Christie Kerr, from the Department of Psychology at the University of Glasgow, shrunk high-resolution passport-style photos and extracted images of bystanders' faces from the reflections in the subjects' eyes.

Although the resolution of the captured images was quite low (some of them were only 27 pixels wide), observers were still able to accurately identify who the bystanders were.

When observers were shown photos to compare faces, they were able to identify strangers' faces 71% of the time and familiar faces 84% ​​of the time.

In another test, observers were able to accurately name a familiar face from a low-resolution eye reflection.

The study shows applicability in solving criminal cases.

For example, analyzing the eye reflections of victims photographed in child sexual abuse or kidnapping cases, or images of people taken from cameras used as investigative evidence, can provide crucial clues about intruders or their accomplices, or their connection to a particular location.

According to Hanoi Moi