henry@utzoo.UUCP (Henry Spencer) (07/03/83)
Positioning a cursor by where your eyeball is looking is certainly
possible (albeit I think it's still expensive and complicated), but
the idea has always struck me as a non-starter for serious work.
Why? Because eye motions are jerky, irregular, and not fully under
conscious control. If you doubt me, check out the existing literature
on how the eye moves. Your brain does an unbelievable amount of
processing on the raw input to give you the illusion of a stationary
field of view and more-or-less smooth eye movement.
On the other hand, positioning a cursor by head movement is quite
feasible and doesn't share these problems. Hardware for this does
exist -- THIS is where the military money is going, with experimental
systems under test in several places -- although at the moment it's
expensive and requires too much gear on and near the head. On the
other hand, the military systems are designed for looking out of a
cockpit at the sky. Using this stuff as a cursor positioner should
be easier, because you can have equipment near the point being aimed
at (i.e. around the screen). Also, the military have to project an
aiming crosshair or something into the field of view, while we can
simply draw a suitable symbol on the screen when necessary. Anybody
know whether there is anyone seriously exploring this?
(If you want to look up the military work -- it shouldn't be hard to
find, there's been a lot of interest in it -- the magic buzzword to
look for is "helmet-mounted sights".)
--
Henry Spencer
U of Toronto
{allegra,ihnp4,linus,decvax}!utzoo!henry