When the XBox Kinect hit the market last November, the web was inundated with a wave of cute little girls scratching tiger-ears, dancing twenty-somethings, and other well-polished marketing material. The hype was met with modestly successful sales numbers, and quite a few of the devices found their way into homes. In spite of the sales and positive reviews, I still kinda felt like many of you do: that the Kinect was just a gimmick that would be forgotten within the year. However, when Microsoft announced the release of a software development kit (SDK) for the Kinect hardware last week I decided it was time to give the Kinect another look. I was amazed that, once past the Kinectimals and the generic fitness games, I actually found a piece of technology that was quiet impressive.
How it works
The hardware technology behind the Kinect isn’t terribly complicated. Using an IR (infrared) light emitter, the device floods the room in non-visible light and then uses a specialized IR camera to measure the light. This camera is then able to measure the distance between objects in the room based on the intensity of the reflected invisible light. The Kinect also uses a standard web-cam style camera to track color, facial recognition, and some basic planar motion.
Kinect gets more intense on the software side however, which analyzes the 3D images captured by the two cameras and compares them with mountains of data about the ways the human body moves and interacts with its surroundings. This data, compiled from millions of samples collected by Microsoft while they were developing the program, is able to target and follow 42 points on the human body, so it can track your hand as it waves . . . or swings a light saber.
Why it matters
While the technology behind the Kinect may be a knockout in gaming, its potential really goes far beyond that of virtual bowling. Over the next several years, I think we can expect this type of technology to start popping up all over the place, from remote-less TVs (finally no more tearing up the couch cushions) to room lights that turn on or off with a wave of the hand! It also opens new doors for the field of robotics. Check out this video in which a computer learns to visually recognize different objects by “looking” at them through the Kinect hardware:
It may not seem like much at first glace, but this is the technology we’ll need if we ever want those robot-butlers people are always harping on about. Using the Kinect technology, a robot could successfully look through the fridge and identify which can of soda you’re asking for, which shirt you want ironed, or which part of your back you need massaged! Robo-servants aside, this technology could also be used to pilot tanks, aircraft, and, most likely, giant brightly-painted robots:
Imagine conducting a war overseas fought entirely by machines controlled remotely by technology not so different from the what you’re using to play “Dance Paradise!”
These possibilities are the reason why the release of the Kinect SDK is news-worthy. Microsoft is allowing hackers, researchers, and others free access to the techy innards of the Kinect, leaving them free to modify, alter, and expand the uses of the device, thereby sowing seeds for hundreds of new technologies and developments in the future! To play you out, here’s a video of a guy using a hacked Kinect to control Windows 7: