There are 2 ways of working with kinect in processing:
1. the OpenKinect (libfreenect) drivers, which are 100% FOSS and give you the image and 3D point data
2. OpenNI/NITE, which is free but only partly open source, and includes skeleton tracking and hand gestures
there is also the Microsoft Kinect SDK, which is similar to OpenNI, but only runs on Windows.
Simple OpenNI: an OpenNI/NITE wrapper for Processing
http://code.google.com/p/simple-openni/
Using OpenKinect with Processing:
http://www.shiffman.net/p5/kinect/
You can also combine either of these with OpenCV to use additional computer vision techniques with the RGB or depth data: