Whether it’s a robot working in a disaster area, a self-driving car running around a city, a satellite looking down through space, it’s very useful to have a machine that can see through clouds, fog, fog. Scientists may have made the best system yet.
The newly developed system works through algorithms that measure the motion of individual light particles or photons that are fired in rapid pulses from a laser, and use them to reconstruct objects that are obscured or hidden by the human eye.
What makes this technique even more special is how it can reconstruct the existing light. scattered Bounced off as a barrier blocking the road.
In the experiment, the laser sight was able to see objects hidden behind an inch of foam layer.
“Many imaging techniques make the image a little nicer and a little less noisy, but this is what makes us see what we can’t see.” Says electrical engineer Gordon Wetzstein., At Stanford University.
“This really pushes the limits of what can be done with any kind of sensing system. It’s like a superhuman vision.”
When a laser beam passes through a barrier (foam), in this study, only a few photons hit the object behind it, and fewer photons get the object back. But the algorithm is smart enough to use these little pieces of information to reconstruct hidden objects.
Officially known as confocal diffusion tomography, it’s not the first way to get through barriers like this, but it does offer some improvements. For example, it can work without knowing how far the hidden object is.
This system, like other approaches, can work without relying on ballistic photons. These photons can travel to and from hidden objects through the scattering field, but do not distort themselves.
“We were interested in collecting all the scattered photons to create an image and reconstruct the image through the scattering medium without these assumptions.” Electrical engineer David Lindell says., At Stanford University.
“This makes our system particularly useful for large-scale applications with few ballistic photons.”
Large-scale applications, such as navigating autonomous cars during heavy rain or capturing images of the Earth’s surface (or other planets) through cloud fog-there are many potential uses here. Researchers are continuing to experiment with more scenarios and more spawning environments.
The current system does not deal particularly well with light scattering caused by fog and haze.
For example, LiDAR is excellent at detecting objects that the human eye cannot see, but when rain or fog interferes with detailed laser scans, problems start to arise. Further, this system can solve that problem.
Before going ahead, scans using this method can take anywhere from a minute to an hour, so there are still plenty of optimizations to work with.
That said, it is a very impressive achievement to reproduce hidden objects in three dimensions that cannot be seen by the human eye.
“We are excited to be able to further develop this with other types of scattering geometry.” Lindell says.
“So it’s like seeing objects hidden behind thick material slabs as well as objects embedded in dense material surrounded by fog.”
The research results are Nature Communication.