Many people will be aware of the ‘uncanny valley’ effect, where something that is meant to look human looks ‘off’, becoming slightly sinister. This effect happens paradoxically in computer graphics as they become more sophisticated, creating faces that are almost, but not quite human enough to pass. How can this problem be overcome? One solution is to simply scan humans into virtual reality (VR).
Of course, it isn’t as simple as that. Human bodies are complex and varied. Current VR applications usually use some manner of animated avatar to represent the players, avoiding the uncanny valley effect by using highly stylised art styles, or simply covering the face in a mask or helmet. With VR entering into the social media area, however, the need for realistic human bodies and faces in VR is becoming apparent. Scientists at the Fraunhofer Heinrich Herts Institute (HHI) in Berlin have developed a VR scanning technique to capture a realistic image of a person and transpose them into a VR environment.
The method uses industrial stereo cameras, which use two lenses to capture the image. An array of 32 stereo cameras is set up, allowing for full 360-degree capture of movement.
“We film the person directly in 3D,” Ingo Feldmann, head of the Immersive Media and Communication group at Fraunhofer HHI, said in an interview with Digital Trends, “in what we call volumetric video. There are cameras arranged around this actor … and these cameras create 2D images. From these images we create a 3D structure or model that looks completely natural. There are no animations, it’s the real person, and they can be inserted into a VR or AR experience.”
At the moment, the technique captures an image that is far too large and complex for most VR software to handle, but as technology develops, it might be possible to scan yourself right into VR using the HHI technology.
VRFocus will continue to report on new developments in VR technology.