IDs: Recovering Faces in Eyes Reflections

IDs: Recovering Faces in Eyes Reflections
1 Star2 Stars3 Stars4 Stars5 Stars (1 votes, average: 5.00 out of 5)

Rob Jenkins and Christie Kerr respectively from the Department of Psychology, University of York, UK and the School of Psychology, University of Glasgow, UK, published an interesting research paper in PLoS One last December.

Often seen in fiction series and movie tech dramas, the ability of identifying a human face from a photo in the reflections of the eye of a person turned to be a real possibility with an accuracy rate above 71%.

The experiment used two group of volunteers where one was familiar with the target subjects while the other not. High resolution photos were taken of the subjects with bystanders in their line of sight. The researchers then zoomed in the photos and recovered poor quality images from the subjects reflections.

The two group of volunteers were then given the task to face-match these poor quality photos with normal photos of the same subjects. The success rate for the group unfamiliar with these faces reached 71% while the one for the familiar group reached 84%.

“The pupil of the eye is like a black mirror. To enhance the image, you have to zoom in and adjust the contrast,” says Rob Jenkins of the department of psychology at the University of York.

The paper concludes the research with interesting possibilities:

“One possible extension of this technique would be to combine pairs of images recovered from the subject’s two eyes. In principle, these images contain the stereo disparity information required to reconstruct a 3D representation of the environment from the viewpoint the photographic subject. Since corneal reflections extend beyond the aperture of the pupil, such reconstructions could capture a wider angle of the scene than was visible to the subject at the time.

For now, our findings suggest a novel application of high-resolution photography: for crimes in which victims are photographed, corneal image analysis could be useful for identifying perpetrators. As with other sources of forensic evidence (e.g. fingerprints), corneal reflection images may not always be readily available. In particular, clear corneal reflections require the subject’s face to be in focus, and viewed from a roughly frontal angle under good lighting. They also require high image resolution in order for bystanders’ faces to be properly resolved.

We note that pixel count per dollar for digital cameras has been doubling approximately every twelve months. This trajectory implies that mobile phones could soon carry >39 megapixel cameras routinely. However, as the current study emphasizes, the extracted face images need not be of high quality in order to be identifiable. For this reason, obtaining optimal viewers – those who are familiar with the faces concerned – may be more important than obtaining optimal images.”

Sources: / PLoS One
Photos / Video Credits: PLoS One – From the article.