It is a given that the Kinect can detect multiple users at once but this Kinect hack teaches the community the possibility of user isolation. Ben Gotow showcases this video on how the Kinect is able to pick out only one user from several detected by the Kinect and also produce the programmed effect to that user alone. In this video, the effect is the distortion of the user’s image. Given that there are 3 people using the Kinect, the program is able to detect the figures and select only one to give the distortion effect. The output is an interesting interaction between 2 normal looking figures and one distorted by the Kinect. The great thing about this Kinect hack is that despite having multiple users, the Kinect community now knows that it can program their hacks to respond to only one user or have their video effects on only select users. This means that gesture-based programs or Kinect video effects can be further enhanced.
The program made use of GLSL Shaders.
Here is a description by the developer:
“Early working prototype of my kinect hack which distorts an individual user in the scene. It uses the PrimeSense framework to do scene analysis, extracts an individual blob from the scene, separates it from the rest of the image and fills the background with cached “deeper” RGB data and then uses a multitexturing GLSL shader to distort the user and draw them back in over the background.”
For more information about the Kinect User Isolation and Distortion, visit the project’s Vimeo Website.