On Tue, Jul 30, 2013 at 9:13 PM, Johannes Scholz <[email protected]> wrote: > I work in a Virtual Reality environment where the ball-marker-based infrared > ART Tracking is still state of the art since years. It is very robust and > reliable while delivering good precision, altought not on a sub-mm level. Now > on the other hand its very expensive, we're talking about many 1000 Euros. > The Leap costs 1-2% of that solution, so some kind of drawback might be > acceptable. > > But....I think although the Leap is not perfect, you can make great stuff > when using the Leap correctly. Finger tracking is not that reliable, but hand > tracking is quite good I think. After some training I can now use the > LeapManipulator to navigate through my models reliably. But it can do more... > I will try to make some demos of what I have in mind in the next weeks.
I am also in VR field and the main beef I have with Leap is: a) tiny workspace - every larger gesture makes you exit the workspace, losing the tracking - a very jarring event. It is a pity that they have made the device so tiny, compromising the useful working volume in the process. b) the hand orientation is sketchy at best I have tried your manipulator code and while I can translate the model just fine, I rarely succeed to reliably rotate or scale it - mainly because I always run out of the tracked workspace. So basically that leaves hand position detection +- some fingers, in a tiny workspace and typically requiring a tiring pose to use. I had much better success with something like the Razer Hydra for low cost desktop applications and when I need more space (e.g. working standing up), the Kinect. A similar device from NaturalPoint - the TrackIR - works a lot more reliably than the Leap. You don't get to wave your fingers at the screen so much, but the software is a whole lot more robust. The Leap is a solution looking for a problem, in my opinion, especially with the buggy software they have. Jan _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
