A pair of bioengineers and an academic at UCLA have devised a new 3D scanning camera system that is significantly more advanced than the current state-of-the-art and can figure out the size and shape of objects hidden round corners or behind other objects. The technology could find its initial home in autonomous vehicles and medical imaging.
The idea for the cameras came from two natural phenomena found in flies, who have visual systems that can scan through nearly 360 degrees but only have a fixed focal length, and bats, who use echolocation or sonar to navigate in the dark.
Their novel CLIP (Compact Light-field Photography) imaging framework “sees” with extended depth range and around objects using simple optics and a small array of sensors. An array of seven LIDAR camera take a lower-resolution image of a scene, processes what individual cameras see, then reconstructs the combined scene as a high-resolution 3D image.
Source: Engineering & Technology