![](bg2.png)
2.1 Augmented Reality
The term Augmented Reality was developed in 1990
by Thomas Caudell to refer to a system used in Boe-
ing to assist the cable assembly of aircraft. In 1997
Azuma (Azuma, 1997) gave the first definition of aug-
mented reality as a system that:
• combines reality with virtual elements,
• is interactive, and
• renders 3D objects.
One of the most important aspects to be consid-
ered in the design and implementation of an aug-
mented reality system is the environment that it is go-
ing to represent. This consideration will determine its
features and complexity. In the literature, several sys-
tems have been developed that manage both indoor
and outdoor environments. Indoor environments are
much less restrictive, allowing the use of more pow-
erful computers and fixed systems based on a previous
training of the environment. In outdoor environments,
the user has to transport the entire system, which lim-
its the available processing capacity. Moreover, the
impossibility of preparing or controlling a hostile en-
vironment, subjected to extreme magnetic phenom-
ena, light or weather changes and other natural phe-
nomena appears.
Depending on the environment that is represented
and the future application, an augmented reality sys-
tem is composed of the following elements:
• A processing system.
• A visualization device.
• A monitoring system.
The processing system can be a fixed or portable
computer, PDA or mobile phone. The choice depends
on the work environment and the process capability
required by the application.
The display device is the part of the system re-
sponsible to render the augmentation. The potential
devices can be divided into two groups: the ones
based on optical technologies and the ones based on
video. On one hand, in the systems based on optical
technologies, the user directly observes the real scene,
overlapped by the synthetic image by means of opti-
cal combiners. These devices are usually mounted on
the user’s head, such as virtual reality helmets. They
usually present some problems such as ghosting, eye
strain or fatigue. On the other hand, the video-based
systems, by contrast, combine a video sequence from
the actual scene, captured with one or more cameras,
with synthetic images by mixing video techniques.
The advantages of this kind of systems are their sim-
plicity and cost, since they only require a personal
computer and some USB or Firewire video cameras.
Furthermore, the availability of a sequence of real-
world video allows us to use tracking systems based
on the detection of characteristic traits using computer
vision techniques.
Finally, the tracking system is in charge of estimat-
ing the position and orientation of the the real-world
view to augment. The type of system to be used is
determined by the environment in which the applica-
tion is going to be applied and by the utilized display
system. In an optical-technology-based system, some
sensors have to be used to implement the monitoring
system. However, in a video-based system can be ap-
plied computer vision techniques that complement the
use of sensors, such as GPS receivers, magnetic sen-
sors (bars), inertial sensors (accelerometers and gyro-
scopes) among others.
In recent years, some augmented reality systems
based on video have been proposed that perform the
monitoring by detecting characteristics or features of
the image. The insertion of markers in known posi-
tions in the real environment facilitates the monitor-
ing, but requires prior preparation of it.
Since the initial proposals at the beginning of the
90’s, augmented reality systems have been applied in
fields such as medical visualization, repair, assembly
and maintenance of machinery, planning of actions
for robots, entertainment, education or construction,
among others (Azuma, 1997). Most of these appli-
cations require very complex and expensive systems
and they usually operate in interior environments. Al-
ternatively, in recent years, some applications have
been implemented that use a personal computer or
mobile device, a webcam and a set of low cost sen-
sors. Among them, we highlight the animated base-
ball cards for sale in the United States, or Wikitude
World Browser for the iPhone (Wikitude, 2011).
Many of the proposals about low cost applications
use open source libraries, such as ARToolKit (AR-
ToolKit, 2011). ARToolKit includes a tracking sys-
tem based on markers that allows us to show three-
dimensional objects superimposed on the real image
captured by a camera. Its main advantages are the
availability of the code and the few requirements on
devices (a computer and a video camera), so it be-
comes an ideal platform for developing augmented
reality applications. Its main drawback is the need
to prepare the work environment by placing mark-
ers, which makes it difficult to use in outdoor envi-
ronments. However, this library has been used in nu-
merous applications indoors (Kwon and Park, 2005)
(Asai et al., 2004) (Nischelwitzer et al., 2007), and in
some outdoor applications combined with other mon-
itoring techniques (Guo et al., 2008).
GRAPP 2012 - International Conference on Computer Graphics Theory and Applications
454