Many of us take for granted our 3D vision, which comes from having two eyes and a brain able to combine images from both. Those who are attempting to give 3D vision to machines however know how hard it can be to achieve. Now researchers have found a way to greatly improve one means of generating 3D images, as reported in the Optical Society's journal, Optics Express.
Instead of using two cameras and combining a pair of images, many machines use a time-of-flight method to gather 3D information. This approach measures how long it takes a photon to reflect off of an object and return to a detector, but is plagued by issues including short range. These researchers overcame that issue though by using infrared photons and a detector for counting individual photons. Because infrared light is not disrupted by the atmosphere as much as visible light, this system can be used to collect 3D information on objects as far as a kilometer away, while still being high-resolution, thanks to the precision of the detector.
Unfortunately the scanner does have some flaws at the moment, including being relatively slow at processing the information which is collected in just seconds. Currently it takes five to six minutes to process the data, but the researchers believe they can cut this down in the short term with a more powerful computer and in the long term with processors dedicated to this task.
Source: Optical Society of America