From: PSOTKA%STARS.decnet@alexandria-emh2.army.mil
Subject: SCI:  Tracker use for head and shoulders:  Ideas?
Date: Fri, 11 Jun 1993 19:20:31 CDT


VIRTU-L


I had some intersting discussions recently with Sandy Ressler and Rob
Jacobs about how to use separate trackers for head and shoulders
reasonably.

It seems obvious that our "Immersion porcessing" system doesn't much
care whether we move our eyes by moving our head or by moving our
shoulders (body).  The world remains pretty stationary both ways and
we remain immersed in our environments quite easily.  Most trackers
are head-mounted so it doesn't matter whether we move our head or our
shoulders (and thus our head) to get the effects of our moving point
of view.  But is there anything reasonable we could do with the
information that it was our body/shoulders that moved rather than just
our head?

One idea we had was that moving your body should result in a
"re-centering" of the display.  So for intance, if we were looking at
a map and moved oout head to the right, the view should shift to the
right too; but then, if we moved our body to the right, the
displayshould compensate to make that new point of view straight
ahead.  That would let you swing your head back left while the display
remained stationary (as if you had moved your chair over to the right.

Using the same sort of technique, imagine using an augmented reality
display of your workstation at your desk.  Someone comes in and you
swing your chair to start talking to him.  Wouldn't it be nice if your
workstation swung with your body(shoulders) but you could still peruse
the whole display with headmovements.  Perhaps you would not want the
display to move constantly with every body motion, but instead track
your body movements only slowly, with a delay that looked to see if
your body had swung more or less permannetly to a new position?

Any other ideas about taking advantage of separate head and shoulder
tracking ?

joe