Revision history [back]

OpenCV is a computer vision based library. If you are interested in machine Learning, you should consider using a dedicated library for that. There is a lot, depending on your languages (libsvm, mlpy, etc.), and some of them are integrated into OpenCV, but I don't see the point to used OpenCV. Maybe I miss something?

At least, you could try to transpose your motion into an image (like if you have used pen) and then compute the recognition with OpenCV, but I don't think it would be the best solution...