Yesterday I bought a Kinect and hooked it up to my PC. After a couple of hours of compiling drivers and finally realizing I had to switch USB ports for the camera to work I could start writing code.

I used the CLNUIDeviceTest as a base since it already had the lines of code required to pull the image from Kinect. My initial problem was how to access the raw bytestream of image data. I overcame this and made a small copy operation that “makes it appear as if you are walking through a wall”. At least that was the point, but the Kinect obviously needs some calibrating and I’m copying source image wrong. Both easy to fix.

Anyhow I just wanted to share the code used for raw image access. Speed is great


And a sample picture…


As you can see the black region on the right represents the picture while the blue region represents the height. I’m guessing I’m sitting a bit too close, maybe I should be processing the green channel instead.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.