javascriptjquerykinectopenkinectkinect-interaction

Controlling web browser JS components with Microsoft Kinect


I am planning to start a personal project involving controlling jQuery components such as a carousel with Microsoft Kinect. Only basic gestures are required such as swipe left/right, stop, click.

I have been digging around for answers here and there, and I've come across a few Kinect implementations/frameworks which could be used for the project:

Kinesis.io https://github.com/Kinesis-io

DepthJS http://depthjs.media.mit.edu

KinectJS http://kinect.childnodes.com/

OpenKinect http://openkinect.org

I am after some advice on what would be the best way to tackle this project is, as getting the initial design/architecture right is probably more than half the task. The more responsive/fast it is, the better, and the simpler it is, the better. It only needs to work for Kinect v1 (unless there is a better solution for v2?) and there are no requirements on the OS however I would prefer if it works with OS X.

Perhaps just going Windows and enabling OS level gesture control might work better and easier? But then again, if I can't capture the controls at a JS level, I may not get the level of control I need. Just a thought.

Any advice would be appreciated, thanks!


Solution

  • Did a similar POC in 2011 and even though there are slot of great libs around now I think the key here is to use websockets.

    I would not focus on jQuery components but instead make sure that communication between the kinect and the browser is working, what you do with the commands in the browser is secondary.

    IMO using windows with the kinect will be easier than using OSX and sending data to JS is still not a problem.