A team of University of Wisconsin researchers have received a major grant from the Department of Defense to further develop a sophisticated piece of optical technology: A camera that can see around corners.
The idea behind the groundbreaking tech is that light particles from a bright flash can be collected by camera sensors and analyzed to visualize objects hidden from view. Through recent experiments at the UW, researchers have borne out the theory.
The team used a laser to fire a pulse of light at an angle toward a far wall, resulting in some of the light particles reflecting off the surface and bouncing across objects set up around a corner. Those particles eventually bounce all the way back to near the point of origin, where they're collected by a camera.
"The light comes back, like an echo essentially. And based on what that light looks like, you can reconstruct an object using a computer," said the imaging specialist Andreas Velten.
Velten, a researcher with the Computational Optics Group Laboratory for Optical and Computational Instrumentation and the Morgridge Research Institute, is largely responsible for pioneering the "scattered-light" imaging system.
Velten's field specializes in using computer processing to take imaging beyond simple, visual recreations of what a person could otherwise be able to see — in other words, beyond photography in the traditional sense.
"We want to design imaging systems that allow us to capture information that's beyond the human eye," he said.
The scattered-light imaging technology is a significant example of such a system. It provides a possible answer to an old optical imaging problem — how to get information about environments obstructed from view — and comes with a wide array of potential applications. According to Velten, the technology could be used whenever "you need to know what's in a room, and you don't have the resources, or it's too dangerous, to put someone in that room."
During natural disasters, like an earthquake, the camera could be deployed to scan rooms for survivors in an unstable apartment building, said Velten. On an assembly line, if a manufacturer wants to examine the inside of a machine with lots of moving parts, the camera would also come in handy.
The applications extend to robotics as well. Velten said that this kind of technology could make a huge impact in the effectiveness of self-driving cars, for example. He added that NASA had even talked with his team to discuss the potential of the technology to help researchers map out tricky-to-explore caves on the moon.
At this point, said Velten, it's clear that the theory and technology works. The $4.4 million he and his research team will receive from the Defense Advanced Research Projects Agency will now go toward further expanding upon the potential of the idea.
"In order to make this a feasible technology, we need to make a lot of progress," said Velten.
Part of the team will work on the engineering and design of the technology. Right now, said Velten, the lab equipment they're using doesn't translate into something that could be deployable in the real world. The hardware will need to become more compact and light — eventually, with the idea being that the technology could become integrated into a smartphone.
On top of that, said Velten, there are other theoretical questions about the technology that need to be explored.
"What are the limitations of this? What can we actually do?" said Velten.
They've solved the problem of visualizing objects that are around a corner, said Velten. Now, he said, the question becomes: What about two corners?