Kinect Programming

There aren't any tutorials at the moment -- but I can write one if that would help everyone to get started. What would you like to see in a tutorial, beyond the code that's in the painting example? For example, are you interested in doing things with stick figures on the screen, or using the Kinect for interesting control interfaces, or using the image information to do cut-outs of people? I do have several more scenarios that use the Kinect. They're not quite as easy to follow as the painting one, but we could maybe do another video and upload the source for those, too.

I personally am looking at using the kinect to output a set of commands to be sent to the Lego Mindstorm robots. But in general I think it would be good to know how to even start acquiring the data and how you actually got greenfoot to track the user. I think a good tutorial would be for the painting one in how you got it to work. But then maybe add the ability to change colors on the fly, or have multiple people. i personally have no idea on how the coding works for getting the data from the kinect. So something along those lines so that you can give the information to more people showing them how to aquire the data and then use it. I know a lot of it depends on what the person wants to do with the kinect, but it would be good to know how to get started with the kinect.

How I got it to work is not something that I want to write about (short answer: C++ and sockets). The key thing that you need to know for using it in Greenfoot is what the API is in Java: there is the getTrackedUsers() method in KinectWorld which gives back the list of UserData. From there you can getAllJoints and use the joint positions. The idea is that you get started by customising the painting scenario to do what you want, rather than somehow starting from scratch and building your own Kinect interface :-) I will see if I can put together a tutorial/guide on the main data types in the Greenfoot Kinect API.