Biomedical Engineering Reference
In-Depth Information
users must essentially recalibrate their movements to account for the movement of
the surface under their feet.
As an alternative to these hardware-based solutions, several researchers have
undertaken the development of software-based systems to allow physical naviga-
tion through VEs that are larger than the available physical space. Such techniques
include combinations of redirected walking [
81
], motion compression (e.g., [
71
]),
and reorienting users who approach the space's limits (e.g., [
74
]). These techniques,
which we will refer to in aggregate as redirected walking, work by subtly altering the
virtual display in an effort to induce a change in the user's movements. For example,
by slowly rotating the VE leftward about a user's viewpoint while she is attempting to
walk forward, shewill—without realizing it—adjust her course to the left.When done
correctly, a user can be induced essentially to walk in circles and thus to use a rela-
tively small physical space in navigating a much larger virtual one. Again, these tech-
niques introduce non-veridical idiothetic sensory input that conflicts with the visual
and auditory sensory input that users perceive regarding their movements. However,
these conflicts are typically designed to be below consciously detectable thresholds
and seem not to impact spatial learning [
45
] or increase motion sickness [
100
].
Uses for HMD Systems
Like other VE systems, there are situations for which an HMD-based system may
be more or less appropriate. If it is important for users to be able to navigate by
means of natural ambulation—for example, when familiarizing users with a large-
scale environment—then an un-tethered HMD would be particularly appropriate.
Additionally, HMDs are ideal for occluding users' view of the surrounding environ-
ment, thereby removing visual distractions and increasing the sense of immersion.
For example, if a user is seated on a motion platform (e.g., [
9
]), it may be ideal to
remove it from view so that the resulting motion appears to be coming from a virtual
source and not a visible physical source.
Conversely, HMDs are not ideal for situations in which it would be advantageous
for users to see their own hands, feet, or body within the VE—an inherent feature
of CAVEs. Similarly, multiple users can gather around a desktop VE or step into the
same CAVE to share a virtual space. Because a user wearing an HMD is essentially
blindfolded to the physical world, any additional users and the user's own body must
be simulated within the VE, which can require substantial additional resources in
terms of motion tracking, communication, and computational load. Finally, HMDs
are adequate but perhaps not ideal at rendering close booth-sized environments that
require little or no navigation, such as a cockpit or driving simulator. As mentioned
above, the visual expanse of a CAVE makes it very well suited to these situations.
Because an HMD typically restricts the user's FOV, the same scene will require
substantially more head movement to apprehend in an HMD relative to a CAVE,
which probably adversely affects the quality of acquired spatial knowledge [
27
].
