Game Development Reference
developed for virtual environments. There are also pointers that help to select an area
of the screen, like the “wand'' that consists of a location tracker (generally electro-
magnetic) and several binary command buttons. According to the orientation of the
tracker, a cursor is displayed at the intersection of the direction of the pointer and the
screen plane. We can also have another type of pointer by adapting command inter-
faces developed for multimedia. For example, we can use “interactive'' laser pointers
in virtual reality. Technically, a user can point to a small area on a large screen using
a hand-held laser beam (Olsen & Nielsen, 2001). If the screen is filmed by a camera,
the computer can, in real time, know the virtual object or the menu indicated and thus
provide a response. It is thus a selection tool on a 2D plan that can be adapted in a
VR application. The Barco company offers the “Magic-Y'' laser pointer, which works
with infrared cameras (in visible light, the laser beam would have to be very intense,
and hence dangerous, for it to be seen in the light rays of the video projectors).
Other command interfaces are based on tracking the movements of the subject
using a camera. For example, the Alterface company uses a camera that analyses
gestures to make it possible for the user to navigate in a virtual environment.
7.3.1 3D Mouse
We can define a 3D mouse as an interface that helps the user to command, in general,
three perpendicular translations and three perpendicular rotations in a space. As has
already been shown, please remember that a tracker can perform the same function,
but a 3D mouse cannot indicate the absolute location of the hand. To the advantage
of the 3D mouse, it is technically easier to produce these functions by playing on the
voluntary hand movements. This generally results in a relatively low purchase price
with respect to the location sensors.
Please note that although it is not too difficult to translate any virtual object
(cursor, etc.) in any direction using the three translation commands, the orientation of
the cursor is sensitive because the rotations following three perpendicular axes are not
commutative geometrical operations. The orientation of a virtual object in a desired
direction (and not at random) can only be done by persons who regularly use this type
of interface, like the engineers who work on CAD work stations everyday.
The 3D mouse can be based on the principle of a 2D joystick to which we can
apply a supplementary vertical force for the third dimension. In this case, we only
control the three degrees of freedom in translation.
Another similar principle is based on the handling of a ball on which we manually
apply measured forces and torques. Generally, these variable efforts proportionately
command the rotation and translation speeds of the object displayed on the screen.
Please note that the ball moves less compared to the base that it is attached to, which
makes it possible for the operator to keep his hand practically still. This type of interface
measures efforts, but only for a movement command, it is thus not used as a force
The technical principle is based on a ball equipped with sensors (three for the
forces and three for torques) measuring the efforts of the operator's hand on a less
deformable element. The forces and the torques are indirectly recorded by optical sen-
sors that measure low movements caused by deformations. These devices are affected
by the coupling of forces. It is difficult to apply forces without torques and vice versa.