So we’ve seen motion capture, which detects 3D movement, and we’ve seen 3D imaging, but scientists at the University of the West of England (yes, that’s its name) have cooked up a way to take three-dimensional moving images of the face in high resolutions. Technology like this could greatly improve security at places like airports, reports NewScientist.
Here’s how it works: a camera captures a bunch of high resolution 2D images of someone’s face as its lit up by pulses of LED lights. Using the shadows and light reflection information in conjunction with the original picture, a computer program can determine skin color and face shape, producing a 3D model of the face that can move and animate in real time.
Prof. Melvyn Smith, who leads the project at the university’s Machine Vision Laboratory, explains: “We believe our technology is the first to capture a moving face at sub-pixel resolution, where the user can interact with the model by altering the pose and lighting in real time.”
The video below should give a good idea of how the technology works. It’s not a full 3D model of the head, but it has enough dimension that facial recognition could be extremely accurate using a technique like this, which he labels “4D.”
- Bored by Skype and FaceTime? 3D video chat is on its way to a phone near you
- Here is the best photo gear to come out of CES 2018
- ModiFace replaces makeup brushes with neural networks, and it’s coming to the S9
- Lenovo’s Mirage Solo headset and VR camera are available for pre-order
- Prototype Insta360 light field camera lets viewers move through the scene