- Hacking the Kinect involves connecting it to a computer instead of an Xbox 360 and using open source drivers to control and receive data.
- The Open Kinect site provides the open source drivers and instructions for controlling and getting data from the depth and color cameras.
- To turn the depth video into something useful, there are two packages: OpenCV and OpenNI/PrimeSense.
- OpenCV has a code sample for integrating with Kinect, while OpenNI/PrimeSense requires following instructions to get them working together.
- The OpenNI/PrimeSense route was the preferred option for the author's purposes.
Hacking the Kinect is as cool as it sounds. It isn’t hacking in the purest sense. It is more like connecting the Kinect to your computer instead of your XBox 360. You use the open source drivers to control and receive data. The kinds of projects being done with the Kinect are very cool.
The Open Source drivers and instructions can be found at the Open Kinect site. These drivers only allow you to control and get the data from the depth and color cameras. Unless you plan to take pictures or video with the Kinect, you have to do more.
There are 2 packages for taking the depth video and turning it into something useful. The first is OpenCV. The code sample for integrating OpenCV and Kinect is here. The second is OpenNI and PrimeSense. PrimeSense is the company that provided the sensing technology to Microsoft for the Kinect. The instructions for getting OpenNI, PrimeSence and OpenKinect working together is here.
For my purposes, the OpenNI/PrimeSense route was what I needed. I suggest you look through the sample applications to see how to interact with the Kinect’s data.