Biomedical Engineering Reference
In-Depth Information
estimated object-motion component
(before recalibration)
estimated object-motion component
(after recalibration)
(a)
(b)
Fig. 4.6
Estimated object motion component (
dotted line
) after self-motion component (
dashed
line
) is factored out from the optic flow field, which includes both the self-motion and object-
motion components (
solid line
). The magnitude of optic flow attributed to self-motion (i.e., the
component that is factored out) is greater in (
b
)than(
a
) due to recalibration of non-visual self-
motion information
flow. To detect such information, people must factor out the influence of self-
motion. The findings of Fajen and Matthis (in press [
8
]) show that both visual and
non-visual self-motion information can be used to factor out the influence of self-
motion, allowing for the perception of passability.
4.7 Extensions of the Affordance-Based Approach
In the shrinking gap example above, it was assumed that there are two obstacles
positioned symmetrically about the locomotor axis and converging toward a common
point at the same speed. We can relax this assumption and assume that there are
multiple obstacles at different depths that will cross the locomotor axis in different
places and at different times. In such situations, there are many possible routes
because each obstacle introduces a choice point at which the actor must decide
whether to pass in front or behind. Further, the decision must be made in a way that
takes into account the physical sizes of the obstacles and the observer's body, as
well as the observer's locomotor capabilities. The source of information identified
in the previous section provides a basis for selecting actions in such situations. In
the same way that such information can be used to perceive
ν
min
for the shrinking
gap, the same information can also be used to perceive
ν
min
for each obstacle in the
scene—that is, how fast one would need to move to pass in front of that obstacle.
In addition, by changing the reference point from the leading edge of the obstacle
to its trailing edge, the same information can also be used to perceive the maximum
speed at which one could move to pass behind each obstacle. Thus, one can perceive
the range of speeds at which not to move (i.e., because doing so would result in a
collision) as well as how fast one would need to move to pass in front of or behind
each obstacle. Such information could be used to decide which route to follow.
Further, this applies regardless of whether there is one, two, or many obstacles in the
environment. Therefore, such information could be used to select routes in arbitrarily
complex environments with multiple moving obstacles.
