I'm preparing to start a kinnect-like project. All I have right now is a webcam. What I want to do is ideally recognize my face and two arms then translate their positions, angles, size to midi/osc messages.
The thing is there are like 4,000 different contributed library Computer Vision options and would like the community to advice me on what would be the simplest or best option for this project.
I program ruby/php just fine, and have done some simple projects with Processing.
Ideally the library is easy to use AND easy to understand AND easy to extend.
