Of course, the code has to take that in account : make orthogonal projections with smooth tolerance... no problem doing that. There is a usb device to control the view in games called Track-IR, and there is no problem to accomodate with 3D positioning and move... so I am sure to have no problem with a great interface to deal with it. (some bip can be played once a move is recognized by the device, so you know if you did a bad rotation and have half a second to do it again or if your move was completed and you can now go on next one). Don't forget that we can fix limits on speeds, angles, and ranges... lots of parameters to suit very nice human hands moves. (I just was thinking about having untypical setting: left-hand/right-hand :) ).
Gal' On Dec 3, 2007 3:41 PM, Kyle Bassett <[EMAIL PROTECTED]> wrote: > I think the largest issue with a motion sensing algorithm would be human > reproducibility. I feel the precision of our movement ability is less than > the precision of the device/possible code. If the code is only designed to > pick up direction and not magnitude, then reliability (of valid person > authenticating correctly) would increase. > > -Kyle _______________________________________________ OpenMoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community