![](bg6.png)
#EVKXKV['XGPV
GZ;QWRWVVJGEWRQPVJGYJKVGVCDNG
$CT'HHGEVQTVQWEJGU6QWEJ5GPUQT
Figure 9: Associate output of virtual sensors with target ac-
tivity event
position data of the ultrasonic 3D tag, the software
can detect the target events using the virtual sensors
and the table defined in Step A to C, as shown in Fig.
10
Human activity events
Output of virtual sensors
(touch state)
3D positions from ultrasonic 3D tag system
Refering table where association of virtual
sensors' output and activity events are defined.
Figure 10: Real time detection and recognition of human
activity by virtual object sensor
4 CONCLUSION
This paper described a method for robustly detecting
human activity in real world and a method for quickly
registering and efficiently recognizing target activity.
The robust detection of human activity is per-
formed by sensorizing objects in real world using an
ultrasonic 3D tagging system, which is a kind of an
ultrasonic location sensor. In order to estimate the
3D position with high accuracy and robustness to oc-
clusion, the authors propose two estimation methods,
one based on a least-squares approach and one based
on RANSAC. The results of experiments conducted
using 48 receivers in the ceiling for a room with di-
mensions of 3.5 × 3.5 × 2.7 m show that it is possible
to improve the accuracy and robustness to occlusion
by increasing the number of ultrasonic receivers and
by adopting a robust estimator such as RANSAC to
estimate the 3D position based on redundant distance
data.
The efficient recognition of human activity in-
volves a method for creating virtual objects using the
ultrasonic 3D tagging system and a stereovision and
a method for virtually sensorizing the created vir-
tual objects interactively on a computer. To verify
the effectiveness of the function, using a stereovi-
sion with ultrasonic 3D tags and interactive software,
the authors registered activity such as ”put a cup on
the desk” and ”staple document” through creating the
simplified 3D shape models of ten objects such as a
TV, a desk, a cup, a chair, a box, and a stapler.
Further development of the system will include re-
finement of the method for measuring the 3D position
with higher accuracy and resolution, and development
of a systematic method for defining and recognizing
human activity based on the tagging data and data
from other sensor systems.
REFERENCES
Fishler, M. and Bolles, R. (1981). Random sample consen-
sus: A paradigm for model fitting with application to
image analysis and automated cartography. Commu-
nication of the ACM, 24:381–395.
Ho, K. (1993). Solution and performance analysis of geolo-
cation by tdoa. IEEE Transaction on Aerospace and
Electronic Systems, 29(4):1311–1322.
Hopper, A., Steggles, P., Ward, A., and Webster, P. (1999).
The anatomy of a context-aware application. In Pro-
ceedings of 5th Annual International Conference Mo-
bile Computing and Networking (Mobicom99), pages
59–68.
Hori, T., Nishida, Y., Kanade, T., and Akiyama, K. (2003).
Improving sampling rate with multiplexed ultrasonic
emitters. In Proceedings of 2003 IEEE International
Conference on Systems, Man and Cybernetics, pages
4522–4527.
Manolakis, D. (1996). Efficient solution and performance
analysis of 3-d position estimation by trilateration.
IEEE Trans. on Aerospace and Electronic Systems,
32(4):1239–1248.
Mizoguchi, H., Sato, T., and Ishikawa, T. (1996). Robotic
office room to support office work by human be-
havior understanding function with networked ma-
chines. IEEE/ASME Transactions on Mechatronics,
1(3):237–244.
Nishida, Y., Aizawa, H., Hori, T., Hoffman, N., Kanade, T.,
and Kakikura, M. (2003). 3-d ultrasonic tagging sys-
tem for observing human activity. In Proceedings of
IEEE International Conference on Intelligent Robots
and Systems, pages 785–791.
Rousseeuw, P. and Leroy, A. (1987). Robust Regression and
Outlier Detection. Wiley, New York.
Shih, S., Minami, M., Morikawa, H., and Aoyama, T.
(2001). An implementation and evaluation of indoor
ultrasonic tracking system. In Proceedings of the 2001
IEICE Domestic General Conference.
ICEIS 2004 - HUMAN-COMPUTER INTERACTION
20