Biomedical Engineering Reference
In-Depth Information
base of the objects that define the gap is resting on the ground surface. Thus, when
an aperture is bounded by objects that are not directly anchored to the ground sur-
face, eyeheight-scaled information is not available but both stride-length-scaled and
head-sway-scaled information are. Likewise, when an aperture is resting on a ground
surface that is higher or lower than the ground surface upon which the observer is
standing, eyeheight-scaled information is not reliable but both stride-length-scaled
and head-sway-scaled information are.
The sufficiency of these two dynamic sources of information was recently tested in
an experiment conducted in an immersive VE [
11
]. The VE was monocularly viewed
through a head-mounted display. Subjects approached and walked through narrow
openings between virtual obstacles, rotating their shoulders as necessary, while head
and shoulder position and orientation were tracked. The task was performed in three
VEs (Post, Tall Post, and Wall) that differed in terms of the availability of eyeheight-
scaled, head-sway-scaled, and stride-length-scaled information.
In the Post condition, the aperture was an opening between a pair of cylindrical
obstacles resting on a textured ground surface (see left column of Fig.
4.3
). As sub-
jects approached the obstacles, all three sources of information were available. In the
Tall Post condition, the ground surface was absent and the cylindrical obstacles on
either side of the aperture spanned the entire visual field from bottom to top (see mid-
dle column of Fig.
4.3
). Both head-sway-scaled and stride-length-scaled information
were available in the Tall Post condition but because the ground plane was absent and
the cylinders had no visible base, eyeheight-scaled information was not available. In
the Wall condition, the aperture was an opening between two untextured walls that
spanned the visual field from bottom to top and from the edge of the aperture to the
periphery (see right column of Fig.
4.3
). The only source of information about aper-
ture size that was available in this condition was stride-length-scaled information,
which is based on the visual angle subtended by the aperture. Eyeheight-scaled infor-
mation was unavailable because the ground surface was absent and head-sway-scaled
information was unavailable because the walls were textureless and extended from
left to right, making it impossible to detect the local optical expansion (i.e.,
φ
in
Fig.
4.2
).
Fig. 4.3
Three VEs used in Fath and Fajen [
11
] and information available in each. EH, HS,
and SL correspond to eyeheight-scaled, head-sway-scaled, and stride-length-scaled information,
respectively
