Scientists just found a way for machines to see through clouds and fog

Whether it be robots working in a disaster area, autonomous cars around the city, or satellites of space through space, machines that can see through clouds, fog and fog are incredibly useful. – And scientists have probably created the best system yet.

The newly developed system works through an algorithm that measures the speed of individual light particles or photons, as fired from the laser in sharp pulses, and they are used to reconstruct those objects Which are obscured or hidden from the human eye.

The way to make the technique extra special is to reconstruct the light that is scattered and surrounded by obstacles along the way.

In experiments, laser vision was able to see objects hidden behind a 1-inch layer of foam.

(Stanford Computational Imaging Lab)

“A lot of imaging techniques make images a little bit better, make a little less noise, but this is actually something where we create invisible scenes,” says Gordon Wettgenstein, an electrical engineer at Stanford University.

“It’s really pushing the limits of what can be possible with any kind of sensing system. It’s like supernatural vision.”

As the laser passes through the light barrier – foam, in this study – only a few photons hit the object from behind, and form back again less frequently. However, the algorithm is smart enough to use those small bits to reconstruct the hidden object.

Officially, it is known as Confocal Diffuse Tomography, and while it is not the first way to look at such constraints, it offers several improvements – it can work without knowing that the hidden How far is the object, for example.

The system is also able to operate without relying on ballistic photons, as other approaches do – these are photons that are able to travel through a scattering region and to a hidden object, but without distorting themselves .

“We were interested in being able to scatter the media without these assumptions and to collect all the scattered photons to recreate the image,” says Stanford University electrical engineer David Lindell.

“This makes our system particularly useful for large-scale applications, where there will be very few ballistic photons.”

Large-scale applications such as navigating a self-driving car in heavy rain, for example, or even capturing surface images of the Earth (or other planets) through cloud haze – have many potential uses here . Researchers want to keep experimenting with more scenarios and more scattered environments.

Current systems are not particularly good at handling scattering of light due to fog and haze.

LiDAR, for example, is excellent at detecting objects that the human eye cannot see, but trouble begins when rain or fog interferes with detailed laser scans. Further down the line, this system can fix that problem.

Before we get ahead of ourselves, it is worth noting that scans using this method can take anywhere from a minute to an hour, so there are a lot of optimizations yet to work.

That said, rebuilding a hidden object into three dimensions that the human eye cannot see is a highly impressive achievement.

“We are excited to pursue this with other types of scattering geometric parts,” says Lindell.

“So, not only objects hidden behind a thick slab of material, but objects that are embedded in a densely scattering material, that would be like seeing an object surrounded by fog.”

The research has been published in Nature communication.


Leave a Reply

Your email address will not be published.