Technological advancements sometimes feel like a perpetual fight against nature itself, with scientists working out how to do things faster, more powerfully, and more efficiently. In short, trying to make objects better than we were meant to be. But sometimes the best solutions are directly inspired by the very nature we’re trying to best.
Take jumping spiders, for instance. The arachnids need very strong vision to avoid predators, hunt prey, and move around. However, with brains the size of poppy-seeds, they don’t have a lot of brain power at their disposal. So their distance calculations have to be very efficient to make up for it. This efficiency is possible thanks to their eye structure. Their eyes feature multiple layers of retinas, (humans only have one retina per eye) that each capture what the spider sees with a slightly different focus. One retinal layer might perceive an object sharply, at the same time as another sees it a bit blurred.
“They see multiple levels of focus at all times,” Emma Alexander, a computer scientist at Northwestern University, explained in a statement. “So, they are always collecting pairs of images. Then, their brains could compare these differences in sharpness to judge distance.”
This ability has inspired Alexander and her colleagues to create SpiderCam, an extremely energy-efficient 3D camera. The team presented their research on June 7at the Computer Vision Foundation’s Conference on Computer Vision and Pattern Recognition in Denver, Colorado.
“We wanted to understand whether we could borrow some of the same principles to create an extremely energy efficient depth sensor that could be used in resource-constrained situations where users don’t have unlimited access to power,” added Alexander, who is also a bio-inspired computer vision expert.
SpiderCam perceives depth the same way that jumping spiders gauge distances before jumping. It captures two versions of the same image, each with subtly different focus settings. Then, a custom algorithm analyses the difference in blurriness between the images, and translates the differences into measurements of depth, in real time. Ultimately, SpiderCam uses less than a watt of power to create real-time 3D maps. By comparison, a standard nightlight uses more energy than one watt.
The algorithm lives in a customizable computer chip enhanced for energy-efficient processing, and this prototype creates depth maps at 32.5 frames per second, consuming just 624 milliwatts of power.
“I’m very interested in settings where you’re very resource constrained and can’t just plug a camera into a wall,” Alexander said. “For example, it could be deployed in field settings with limited power. Separately, I also think it’s particularly exciting for applications like augmented reality where you’re interfacing with the physical world and need to know the locations of objects around you.”
A majority of 3D cameras estimate depth with methods that are good, but can necessitate significant computational power, additional energy, and expensive hardware.
This spider-inspired invention could open the door to a new kind of battery-powered gadgets—such as assistive devices, wearable technologies, robots, and drones—that need to assess their environments. In the future, the team plans to improve the technology and incorporate it into small robots and wearable devices.
The post Jumping spiders inspire wildly efficient 3D camera appeared first on Popular Science.
from Popular Science https://ift.tt/qpbe41I
0 Comments