New camera captures invisible details, sees through the skin of fruit
Microsoft Research teams up with University of Washington to develop a camera that can capture the veins in your hand and other things not visible to the naked eye.
Everyone has a different method for determining if produce at the grocery store is underripe, just right or rotten. There's the knock test, the squeeze test and, of course, the smell test. But if the HyperCam, a new type of camera being developed by the University of Washington and Microsoft Research, gets developed, finding out what's happening beneath the surface of your favorite fruit might be as simple as aiming your mobile phone at it.
The HyperCam uses technology called hyperspectral imaging that gathers images from across the electromagnetic spectrum and combines them into one picture. Regular cameras, on the other hand, shoot only in three bands of the spectrum: red, green and blue.
"When you look at a scene with a naked eye or a normal camera, you're mostly seeing colors. You can say, 'Oh, that's a pair of blue pants,'" lead author Mayank Goel (PDF) said in a statement. Goel is a UW computer science and engineering doctoral student and Microsoft Research graduate fellow. "With a hyperspectral camera, you're looking at the actual material that something is made of. You can see the difference between blue denim and blue cotton."
Hyperspectral imaging isn't new. It's been used in industrial applications and in the space program such as in NASA's Airborne Visible/Infrared Imaging Spectrometer (AVIRIS), which can record images across 224 different bands. What's new here is the fact that the Microsoft/UW partnership could eventually produce a camera that's affordable for regular people to buy.
In a paper presented at a computer conference in Japan, the team said it had developed a camera that would cost about $800 (about £515, AU$1,100), but that it might be possible to develop one that works with a mobile phone camera for just $50 (about £30, AU$70).
"Existing systems are costly and hard to use, so we decided to create an inexpensive hyperspectral camera and explore these uses ourselves," Microsoft researcher Neel Joshi, who collaborated on the project, said in a statement. "After building the camera we just started pointing it at everyday objects -- really anything we could find in our homes and offices -- and we were amazed at all the hidden information it revealed."
Unlike AVIRIS, the HyperCam shoots only across 17 bands, but that's still enough to show some pretty interesting things. For one, it can image what's going on beneath the skins of fruit, as you can see in the avocado example in the video below. In tests, the HyperCam was pointed at 10 different fruits and it was able to predict their ripeness 94 percent of the time, the researchers say.
It's also possible the HyperCam could be used for biometric identification, like the way in which you can use your fingerprint to unlock your phone or iPad. The camera revealed vein and skin texture patterns on peoples' hands, which allowed it to differentiate between the images of hands from 25 people with 99 percent accuracy.
The researchers say that a current challenge facing their affordable hyperspectral camera is that it doesn't perform well in bright light. So next, they'll be focusing on overcoming that problem as well as shrinking the camera down to fit on a mobile phone.