Biomedical Engineering Reference
In-Depth Information
1.4.2.2 CAVE Systems
Classic CAVE Systems
A
CAVE
[
23
] consists of multiple large projected surfaces—traditionally at least three
walls that surround a user, and sometimes a floor, ceiling, or fourth wall—that form
a booth in which one or more users stand while interacting with the VE (Fig.
1.1
b).
CAVE systems differ from traditional desktop VEs in that users do not merely sit
and view a simulation on a small screen, but rather step physically into the simulated
space because the projected surfaces surround them. Such systems thus are able to
engage kinesthetic, vestibular, somatosensory, and efferent systems. For example,
instead of pressing a button to rotate the virtual viewpoint to one's left, a user need
only turn his head or shift his feet. Perhaps most strikingly, however, CAVE displays
are generally much more visually immersive than desktop VEs and can present more
complete optic flow to inform users about their movements. Access to a wide field of
view, and particularly peripheral vision, has been shown to be beneficial for gauging
the speed [
106
] and accuracy of navigation [
2
,
95
]. While typical desktop displays
present a simulated view frustum
1
with a horizontal field of view (HFOV) of around
60
◦
, a CAVE display typically displays more than 180
◦
HFOV with three walls or
a full 360
◦
HFOV with four walls. Likewise, systems with a floor and/or ceiling
will offer significantly improved vertical field of view (FOV). Indeed, a six-walled
CAVE is able to simulate a full field of optical flow perfectly regardless of the user's
facing direction, although a user's actual viewing angle may be artificially limited if
the system requires users to wear shutter glasses. These wide FOVs not only allow
peripheral visual information to be displayed, but enable it to be displayed to a
physically appropriate location on the user's retina. It is well known that reducing
the field of view, for example in an HMD [
2
] or a desktop VE with a physically
matched viewing angle [
5
], can reduce performance in spatial tasks. Indeed, when
normally-sighted users wear a reduced-FOV HMD, they effectively reproduce the
inaccuracies in knowledge of spatial layout that are typical of patients who have
peripheral field loss [
27
].
Valid and realistic body-based sensory information in CAVEs, however, is lim-
ited to head, neck, and eye movements. Users cannot typically walk or travel more
than a few steps in any direction (but see below), placing real limits on the amount
of podokinesthetic sensory information that can be acquired during navigation. As
mentioned above, the lack of incorporation of body-based senses—particularly for
navigational tasks requiring knowledge of orientation changes—can impede learning
and produce less accurate spatial knowledge ([
88
]; see also [
16
]).
1
Desktop VEs are rarely constructed to match the physical viewing angle (the visual extent taken up
by the monitor) and the virtual field of view (the amount of the VE that is visible). A user's physical
viewing angle can vary greatly depending on her distance from the display, and it has been shown
both that users do not notice this discrepancy [
56
] and that a wider field of view is advantageous in
many tasks (e.g., [
5
]).