A nerd pretending to be a software engineer.

Testing the Water on Kinect Applications

Written on

Okay, so I am started planning my final year project this semester, and for this first phase, I am only required to do the planning, the documentation and well, the report stuffs, but, since I don’t have any pictures of the Microsoft Kinect application development, I decided to do some coding on it as well.

Since there are so many possibilities on what the Kinect can do, there are several choices on what kind of application I am going to do, I need to narrow it down. My project is trying to use the Kinect to recognize sign language, and translate them. In other words, besides doing the skeleton tracking, I also need to do some form of gesture recognition, and also, I need to at least know a little bit of the sign language itself. Of course, I still know very little of those things.

First thing to do, of course I need to look up for some samples of the Kinect hack itself, even though it is not doing the same thing, at least it can help me get an idea what the Kinect is capable of. It is not a difficult thing, a little bit of search on any search engine would get me a lot of results.

Here is the one that impressed me the most:

The program can scan the room with the Kinect, and the program will process the input from the Kinect, and create a 3d scene with it. Imagine what it can do if the Kinect’s image sensor have higher resolution. I wonder how they did that. Of course, that is much more sophisticated than what I am trying to do, just by looking at the people involved in the project, you can see how serious that project is.

Here are some of the impressive works:

and this video shows how it can be integrated with other devices from Microsoft:

comments powered by Disqus