Blind corners have long troubled drivers, but they might not pose such a hazard for much longer. Researchers at Northwestern University have developed a new holographic camera technology that can peer around corners by reconstructing scattered light waves, quickly enough to spot fast-moving objects like cars or pedestrians.
When light strikes an object, it scatters, and some of that finds its way to our retinas, or the sensors of a camera, allowing the object to be seen. Of course, that means we can’t see objects behind other objects, or through scattering media like fog or skin. But there might be a way to use the scattering of light off multiple objects to see around corners.
Position a mirror just right, and you can see objects around corners. Even without a mirror, that principle still holds true – it’s just that the secondary object scatters the light too much for us to reconstruct the target. But an emerging technology called non-line-of-sight (NLoS) imaging can do just that.
NLoS systems work by beaming light out, which bounces off a surface, strikes an object and bounces back to the surface, then back to a sensor. Algorithms can then create an image of the object around a corner. As you might expect however, images reconstructed in this way can often be low resolution, or take too long to process.
It gets better as the technique also can work at night and in foggy weather conditions
The researchers on this new study say their new technology improves both of those problems. They call it synthetic wavelength holography, and it works by merging light waves from two lasers into a synthetic light wave, which can then be fired off to produce three-dimensional, “holographic” images of objects around corners or behind other scattering media.
“If you can capture the entire light field of an object in a hologram, then you can reconstruct the object’s three-dimensional shape in its entirety,” says Florian Willomitzer, first author of the study. “We do this holographic imaging around a corner or through scatterers — with synthetic waves instead of normal light waves.”
The team says the system can capture fine details in objects lurking within a large angular field of view, and do so very quickly – within 46 milliseconds. That’s fast enough to see and react to a car or pedestrian approaching from a blind corner, and a vast improvement over other early NLoS systems that needed upwards of an hour to calculate.
“This technique turns walls into mirrors,” says Willomitzer. “It gets better as the technique also can work at night and in foggy weather conditions.”
Not only could the technology eventually allow vehicles to sense unseen hazards, but the team says it could also be used to improve endoscopes, both in industrial and medical contexts. The camera system might not need to flex and bend through pipes or intestines, but could scout ahead by beaming synthetic light waves and watching how they return.
“Our technology will usher in a new wave of imaging capabilities,” says Willomitzer. “Our current sensor prototypes use visible or infrared light, but the principle is universal and could be extended to other wavelengths. For example, the same method could be applied to radio waves for space exploration or underwater acoustic imaging. It can be applied to many areas, and we have only scratched the surface.”
The research was published in the journal Nature Communications.
Source: Northwestern University
