One of the most hazardous conditions for drivers is dense fog. Researchers at Stanford have created a new device that can see through clouds and fog that could be extremely important for future autonomous vehicles. Stanford’s technology uses technology similar to tech found on autonomous cars improved with a highly efficient algorithm developed by the researchers.
The algorithm can reconstruct three-dimensional hidden scenes based on the movement of individual particles of light, known as photons. The system was able to successfully reconstruct shapes that were obscured by a one-inch thick piece of foam. Researcher Gordon Wetzstein says that while many imaging techniques can make images look a little better, the team’s technique can make the invisible visible.
Another potential use for the system the Stanford researchers have developed is in space exploration. The technology could allow satellites to image the surface of the Earth and other planets through a hazy atmosphere or clouds. The system pairs a laser with an extremely sensitive photon detector at the table to record every bit of laser light that hits it. As the laser scans an obstruction such as foam, an occasional photon manages to pass through the foam.
The photons that do pass through the foam hit the object hidden behind it and pass back through the foam to reach the detector. The algorithm used in the system uses software to follow those photons and when they hit the detector to reconstruct the hidden object in 3D.
Researchers admit that this isn’t the first system with the ability to reveal hidden objects through scattering environments, but it does circumvent limitations associated with other techniques. Some of those other systems require knowledge about how far away an object is, the Stanford system does not.