Biomedical Engineering Reference
In-Depth Information
be smaller in older adults than in younger adults, and tends to decrease with age
[ 45 ]. Sekiya and Nagasaki [ 64 ] have shown that the walk ratio is a reliable measure
for evaluating pathological and aging walking patterns. The average step length,
step rate, velocity and walk ratio of healthy male and female subjects is reported in
Table 3.1 . The walk ratio does not differ significantly between males and females. It
is also globally invariant except at low speeds in females [ 63 ].
Knowing this walk ratio for a user, it is possible to automatically estimate his/her
step length and frequency depending on his/her current speed. As a result, it might be
possible to more accurately adapt the trajectory of the camera in real-time [ 70 ], or to
animate avatars simply driven by walking speed [ 12 ]. Conversely, it could be possible
to deduce walking speed using step frequency or step length. Hence, knowing the
frequency of the head oscillations (related to the step frequency) of the user with a
simple webcam, it is possible to compute the corresponding walking speed of the
avatar or of the camera [ 71 ].
It is also widely acknowledged that stride length shows a moderate inverse cor-
relation with the age of the walker (r
=−
0
.
68, P
<
0
.
001) as well as velocity, but
the correlation is not as strong (r
05) [ 34 ]. Hence, the walk ratio
seems to depend on both gender and age, which is an interesting property. In VR the
trajectory of the camera and avatar motions can be customized with only few input
parameters (such as step frequency or gait speed).
In everyday life, we usually walk at our own moderate pace. It is well known that
casual walking involves optimization of the energy costs for moving. Briefly, optimal
human locomotion is achievedwhen peoplemove freely in their own selectedmanner
which could be viewed as a definition of casual walking [ 1 , 2 , 8 , 9 ]. In casual walking
it seems that people always walk at a constant speed over a prescribed distance.
Walking speed has thus been implicitly assumed to be a dominant parameter of
casual walking [ 4 , 42 , 84 ]. Men walk generally faster than women but some authors
have shown that it is mainly true for high speeds [ 64 ] and it also depends on age [ 6 ].
One has also to notice that walking speed decreases with age in both genders [ 6 ].
Casual walking speed starts to decrease during the sixth decade for men and during
the seventh decade for women, as reported in Table 3.2 .
Some authors [ 35 ] have demonstrated that cycle duration is more stable than stride
length and walking speed in casual ground walking.
All this knowledge about casual walking can help developers in VR introduce
variability in camera motion and avatar control thanks to only very little information.
Hence, elderly and young avatars, male and female avatars, would each lead to
different camera motions in the virtual environment in a very simple manner. The
method described in [ 70 ] would be very easy to adapt in order to take this type of
information into account.
Kirtley et al. [ 34 ] have also highlighted significant correlations between step fre-
quency (expressed in steps per minute) and four gait parameters. The regression
equations between these parameters are given in Table 3.3 . Again, this type of in-
formation would help camera controllers to adapt more accurately to the situation.
Such type of knowledge has been used in the past to animate virtual humans with
only few control parameters [ 12 ].
=−
0
.
53, P
<
0
.
Search WWH ::




Custom Search