Most "conventional" cameras, including those built into smartphones and tablets, share a common trait: they only process visible light. That makes sense, since it matches our eyes. But there's another type of camera capable of capturing nearly everything we can't see on our own. The official name for that hardware is a "hyperspectral camera", and they typically cost a fortune. The HyperCam aims to break that trend.

HyperCam: An Affordable Camera That Captures What We Can't See
HyperCam

What Is a Hyperspectral Camera?

If you just want to take a photo, any average smartphone will do. If you need something more professional, you'll need dedicated hardware for superior results. But what if you want an image "outside" the visual spectrum? Most likely, you'd have to step into the territory of hyperspectral images. In very relaxed terms, a hyperspectral camera goes "beyond" what our eyes can see, processing additional wavelengths. The two most relevant drawbacks of any hyperspectral camera are cost and size. Both the budget and the space it requires are usually very large, limiting its application to highly specialized environments.

HyperCam: A Compact Solution

Fortunately, folks at Microsoft Research and the University of Washington want to change that with the project HyperCam. It's a compact camera that can record a total of seventeen different wavelengths in the visible and near-infrared spectrum. The camera compares each special frame with a traditional photo, identifies which ones differ most from the original material, and then creates a composite image by combining those frames. What could we see with the HyperCam's assistance? Two excellent examples are veins under our skin, or the ripeness of fruit.

Right now, the HyperCam design carries a price tag close to $800 (a lab hyperspectral camera costs tens of thousands of dollars), and with enough development, integrating the technology into a smartphone could add just $50 to its overall cost. Of course, this requires the camera to undergo a complex miniaturization process, and it will need solid software to go along with it, but we'd love to see this on the market soon.

Official announcement: