Last week, open source drivers for Kinect were released. Shortly afterwards, we showed you a video of somebody using the motion-sensing device as a means of â€œmulti-touchâ€ input.
Today, we have something much, much cooler. A youtube poster by the name of â€œOkreylosâ€ has used Kinect to generate a 3D model of his room, and himself; showing how impressive the tech behind Kinect really is.
The first video shows how Kinect’s able to generate a 3D model, using a combination of its RGB camera and its beams of infra-red light to measure depth. What gets outputted is a crude 3D model of the room. Although it’s missing a lot of data – anything not directly visible to Kinect’s beams – it’s quite impressive. It’s easy to see the possibility of hooking three of these together to do quick, easy full-scale 3d modelling.
The second video shows how Kinect can be used to quite accuratelyÂ – and impressively – measure the size of objects captured.
It’s still early days, and I’m sure intrepid software geniuses will find even more interesting uses for the device. Its application beyond gaming are fascinating. Seeing what it’s capable of though, one can’t help but ask why the launch games primarily consisted of waggling body parts about.
There’s a third video that tackles the question of using multiple Kinects [thanks to NiteFenix for the tip][Thanks to Noodle for the tip.]
Last Updated: November 15, 2010