3D POINTING IN VIRTUAL REALITY: EXPERIMENTAL STUDY AND VALIDATION

creepytreatmentΤεχνίτη Νοημοσύνη και Ρομποτική

14 Νοε 2013 (πριν από 3 χρόνια και 11 μήνες)

94 εμφανίσεις

3D POINTING IN VIRTUAL REALITY: EXPERIMENTAL STUDY
AND VALIDATION


Authors
: Michele Fiorentino, Antonio E. Uva, Pietro A. Renzulli, Giuseppe Monno


Affiliation:
dDis, Politecnico di Bari, Via Amendola 132, 70100, Bari, Italy.


Contact:

Michele Fiorentino,
m.fiorentino@poliba.it


Area:
Simulation and Virtual Reality in engineering design


Keywords
: Virtual reality, Tracking, Human Computer Interaction deployment

Aims and objectives

One crucial component of Virtual and Augmented Reality (VR/AR) systems is rea
l
-
time
position sensing. Almost every immersive application needs several input devices with 3 or 6
degrees of freedom (DOF). The quality of the whole application heavily relies on the
performance of the trackers used. Precision is just one criterion for t
rackers, others are
scalability, reliability, repeatability, update rates, latency, resolution, working volume, costs
and ergonomics.

Previously used VR tracking systems, like magnetic and acoustic ones, suffered from
drawbacks in precision, resolution an
d repeatability of measurements
[4]
. Mainly due to this
reason, most of VR based applications, such as 3D interactive surface modelling and design,
have been limited to case studies, not conform to a real industrial use
[5]
.

Whilst in the past much of the effort was directed towards tracking error reduction by means of
filtering (such as inertial and position prediction), at the present time, research can be finally
addressed to the direct interpretati
on of the user’s interaction and limitations.

During a normal VR session the users should be able to navigate the scene, interact with the
workspace, perform some actions and modifications according to the desired result, in the most
natural and comfortabl
e manner. Different factors may influence human intention in a VR
environment such as hand’s vibration, fatigue, and limb posture. Our goal is to study those
elements by collecting significant data regarding user’s sessions and to provide the tools for
imp
roving the interactions and optimize the interface. In particular the Spacedesign application,
a VR/AR based conceptual design tool
[3]
, is used as test bed for experimental assessment and
review.


Methodology

For the tests we u
sed an optical infra red tracking system
[1]

consisting of two cameras
connected to a PC running the Open Tracker software
[2]
.

In the first phase of the tests, we aimed to quantify the precision and
repeatability of the system
in a standard office/laboratory. A tracked object is precisely located in different points within
the working volume and position and orientation are recorded. A statistical analysis of the
dataset is carried out. The result are

compared with the producer’s specifications.

In the second phase human factors are taken in account. The tests involve the measurement and
statistical analysis of the involuntary human vibrations in a stationary position. The user is
asked to hold a trac
ked object in a imaginary fixed point. Different postures are experimented
such as sitting on a chair or standing still. The influence of time is also taken into account in
order to determine the user’s tiredness and muscular fatigue.

In the third phase, t
he user traces a sequence of test lines, in various postures, and at different
speeds. The objective is to discover the user’s preferred methods and modalities during the
interaction by computation and statistical analysis of the data (such as velocity an
d
acceleration).

Finally we try to identify the critical aspects of the user interaction in a VR environment,
extract the leading parameters which influence them, and provide the base for the development
in the near future of correlated software tools. Fo
urier and wavelet analysis is carried out on the
data and different filters are implemented and tested for real time computation.

Results

The measured systematic error of the tracking system has been verified to match, in normal
working conditions, the sp
ecifications provided by the vendor. The variation in position and
orientation due to natural involuntary human tremor has been observed and analyzed in the
frequency domain. It varies from user to user but it is characterized by a small frequency
window t
hat is about 9
-
25Hz. The magnitude tends to increase with time span and with
positions which involve fully stretched limbs. It is possible to establish a maximum position
and orientation threshold, useful for the operations which involve precise pointing i
n the
Virtual workspace. An adaptive approach is proposed in order to determine the user’s
personalized filter parameters. The systems analyzes on line the user’s vibration and computes
the threshold accordingly.

Conclusions

The experiments have shown that

commercially available wireless tracking systems can
nowadays provide unmatched precision, disclosing new undiscovered possibilities for the VR
based applications. Human tremor is easily detected by the system and real time auto adaptive
filtering is app
lied. This research is basic and very useful for future works which are based on a
VR interaction system such as 3D VR sketching and surfacing.

References

[1]

ARTtrack1 & DTrack IR Optical Tracking System, Advanced Realtime Tracking GmbH,
www.ar
-
tracking.de.

[2]

G
. Reitmayr and D. Schmalstieg.
Opentracker
-

an open software architecture for
reconfigurable tracking based on XML. In Proc. VR2001, Yokohama, Japan, March
2001.

[3]

M. Fiorentino, R. De Amicis, A. Stork, G. Monno; “Spacedesign: conceptual styling and
design
review in augmented reality”; In Proc. of ISMAR 2002 IEEE and ACM
International Symposium on Mixed and Augmented Reality, Sept. 30
-

Oct. 1, 2002 in
Darmstadt, Germany.

[4]

Meyer, K., Applewhite, H., & Biocca, F., "A Survey of Position Trackers," Presence:
Te
leoperators and Virtual Environments, Vol. 1, No. 2, pp. 173
-
200, 1992.

[5]

N.I. Durlach and A.S. Mavor, Virtual Reality: Scientific and Technological Challenges,
Nat'l Academy Press, Washington, D.C., 1995.