The Xbox Kinect is a popular accessory to the Xbox 360 gaming console, but it’s also been hacked away at by a community of people doing interesting things with the device. You see, the Kinect has 2 cameras that can read depth information, which makes the device much more interesting than a simple webcam. In a way it makes the perfect device to scan in imagery of a physical object for scanning, and that’s exactly what I wanted to try.
I’m no hacker, just someone who likes to fool around with these things, so I was lucky to find a program called KinectToSTL which was built specifically for grabbing files that could be printed on a Makerbot. You can grab the OS X app here and if you have a Kinect make sure it’s plugged in via USB to your computer before opening it up.
The interface is very basic with some controls for deciding how deep, tall, and wide the shot should be. You can also adjust lighting but thats more to help you see details on screen. When you’re ready to grab your shot just click the “exportSTL” button and it will throw a file in the data folder. You can open that STL file in ReplicatorG and it should be ready to print.
It’s a really cool way of easily scanning in objects, particularly faces, and getting a 3d model without much cleanup at all. One thing I want to work on is how to get a full 360 degree image of an object. Right now using a single Kinect limits you to only what the camera can see. Either by combining multiple STL files with joining or using a different method all together I think you could get a full model, but I have yet to get that working. I’ve got a few ideas though and will be sure to blog them once I have it figured out.