Game Development Reference
In-Depth Information
may take a long time if the number of commands is significant and if the vocabulary
recognised by the vocal recognition system is generally limited.
Graphic selection is the most commonly used method. There are two situations:
the use of graphic menus (or “widgets'') or the use of virtual tools. The graphic menu
(menu or “widget4'') is a paradigm that has become a quasi-standard for the control
of 3D applications on a workstation. Unfortunately, the transcription of this con-
cept in virtual environments is not without its problems. Many authors have tried to
translate this paradigm in 3D. Some retain the bidimensional character of the concept
(2D menus/widgets), others extend this concept in 3D (3D menus/widgets). Manag-
ing 2D menus in a 3D virtual environment (see the site 3dwm5 and [van Teylingen
et al., 1997]) raises two problems: a problem of positioning and a problem of selec-
tion. Regarding selection, the task that consists of choosing a command in a menu is
essentially a 1D task. Adding extra dimensions, far from being beneficial, is more of
an inconvenience for the user. In 3D, the selection task gets complicated largely due to
the addition of depth that should be controlled when we want to point to one of the
objects of the list. Using a ray for selecting the commands is an interesting attempt to
eliminate problems related to this superfluous dimension, but it needs certain accuracy
in selection. Another much more effective solution is to display the menu on a physical
plane medium. The selection then comes down to a real 2D operation with the phys-
ical plane improving the selection task considerably (in terms of time and accuracy)
(Lindeman et al., 1999).
Among the possible physical media, we find:
the visualisation system screen: the workbench with its horizontal screen simu-
lating a work surface lends itself well to this type of interface (Forsberg et al.,
1998).
a handheld palette: according to visualisation systems, we are talking about either
a transparent palette for large screen projection systems (Coquillart & Wesche,
1999), or an opaque palette for head-mounted display based systems (Szalavári &
Gervautz, 1997; Poupyrev et al., 1998). In these two cases, the palette is handheld
and its position is recorded in real time. A virtual menu is displayed in 3D space in
such a way that it is superimposed on the palette and follows its movements. This
effect is made possible by the transparency of the palette in case of projections
on large screens. With head-mounted displays, it is sufficient to superimpose the
menu on the palette with the user not seeing the physical world.
Resorting to a physical surface as a 2Dmenu medium is an interesting approach. It
not only facilitates selection but also resolves the problem of positioning (at least in the
last case or the menu is hand-held). We would like to state that Lindeman et al. (1999)
also pointed out that the hand-held menu feature improves the selection performances.
Absence of a physical medium raises the problem of positioning. The positioning
of the menu has a significant influence on its accessibility and on the performance
of selection of commands. The following are the four choices of positioning (this
classification is inspired by the works of Kruijff, 2000):
world-centered positioning is a case with the least constraints wherein the menu is
placed freely in 3D space. The absence of a constraint is appealing but can cause
Search Nedrilad ::




Custom Search