Kinect hacking’s been all the rage since Microsoft released the super-successful motion-sensing camera but researcher Oliver Kreylos has done everyone one better. He’s created an inexpensive 3D virtualization system out of two Kinect sensors. Kreylos had already been working with Kinect to control the way it parses visual data to recognize depth and object placement in 3D space. By working some of his code magic and placing two Kinects next to each other on a 90º angle, he’s been able to combine their visual streams into a virtual 3D model of a specific space. The results of his work are in the YouTube video below.
For some context, Kreylos’ work with $300 worth of equipment approximates the millions of dollars that went into the bullet-time technology that the “Matrix” movies popularized more than a decade ago. This breakthrough has some game design possiblities, too, where it could be used to create levels from real-world geometry.