* Although many of the applications are under development,
only some have advanced beyond proof-of-concept or
prototype stages. In fact many serious practical problems do
exist, two of which we offered solutions to in this paper.
These are:
1- The creation of realistic VE models.
2- The accurate non-intrusive tracking of the user's head and
hands.
* Advanced laser range cameras, such as the two presented in
this paper, can be fast and accurate tools for the creation of
VE models. However, they must be properly calibrated and
the acquired images must be perfectly registered. We have
briefly presented the calibration approach and the
registration technique developed in our laboratory. These
methods have been extensively tested and applied to a
variety of environments and objects and gave excellent
results. The calibration facility and the software tools for
registration are currently being used for several collaborative
projects with industrial partners, such as virtual museums and
rapid prototyping.
* The use of video images (gray-scale or color), which are
registered with the 3-D images, is valuable in adding texture
maps to the geometric models to provide visually realistic
environments. Feature extraction from the video images,
along with their 3-D coordinates computed with
photogrammetric techniques, provide useful information to
assist in the registration of multi-views, particularly in large
and complex environments.
* The accurate tracking of user's head and hands can be
accomplished optically by real-time digital
photogrammetry. Our specially designed photogrammetric
tracking system can currently update the targets positions
and orientations at a rate of 15-20 Hz using Pentium PC and
image processing hardware. This speed is sufficient for only
some applications where the user is not moving too fast. For
many applications, however, a much faster speed (about 3 to
6 times faster) is required so that the user will not experience
problems related to latency. We are currently working on
increasing the tracking speed using prediction methods, such
as Kalman filter, to find the position between system
outputs. Also the ever increasing speed of PCs will
contribute to ultimately achieving the desired tracking rate at
a reasonable cost. On the positive side, the accuracy
achieved by our optical tracking system is an order of
magnitude better than any existing tracking device currently
used for VE. The system also does not require uncomfortable
head or hand gear to be put on by the user.
REFERENCES
Adams, J.A., 1993. Virtual reality is for real. IEEE
Spectrum, 30(10) October, pp. 22-29.
Azuma, R., 1995. A survey of augmented reality. Course
Notes 9: Developing Advanced Virtual Reality,
SIGGRAPH'95, ACM SIGGRAPH, N.Y., N.Y.
Azuma, R. and Bishop, G., 1995. A frequency domain
analysis of head-motion prediction. In: SIGGRAPH'95, L.A.,
Cal., Aug. 6-11, ACM Computer Graphics Proceedings 95,
pp. 401-408.
Baribeau, R., Rioux,M., and Godin,G., 1991. Color
reflectance modeling using a polychromatic laser range
146
sensor. IEEE Trans. Pattern Anal. Machine Intell., 14(2),
pp. 263-269.
Beraldin,J.-A., El-Hakim,S.F., and Cournoyer,L., 1993,
Practical range camera calibration. In: Videometrics II, Proc.
SPIE 2067, pp. 21-31.
Bishop, G. and Fuchs, H., 1992. Research directions in
virtual environments. Computer Graphics, 26(3) August, pp.
153-177.
Blais, F., Lecavalier,M., Domey,J., Boulanger,P., and
Courteau,J., 1992. Application of the BIRIS range sensor for
wood volume measurement. NRC/ERB-1038, 19 pages.
El-Hakim,S.F. and Beraldin,J.-A., 1994. On the
integration of range and intensity data to improve vision-
based three-dimensional measurements. In: Videometrics III,
Proc. SPIE 2350, pp. 306-321.
El-Hakim,S.F., Beraldin,J.-A., and Blais,F., 1995. A
comparative evaluation of the performance of passive and
active 3-D vision systems. In: St. Petersburg Great Lakes
Conference on Digital Photogrammetry and Remote
Sensing, St. Petersburg, Russia, June 25-30, Proc. SPIE
2646, pp. 14-25.
El-Hakim, S.F. and Pizzi,N., 1993. Multicamera vision-
based approach to flexible feature measurement for
inspection and reverse engineering. Optical Engineering,
32(9), pp. 2201-2215.
Goble, J.C., Hinckley,K., Pausch,R., Snell,J.W., and
Kassel,N.F., 1995. Two-handed spatial interface tools for
neurosurgical planning. IEEE Computer, 28(7) July, pp. 20-
26.
Godin, G., Rioux, M., Baribeau, R., 1994. 3-D
registration using range and intensity information.
In:Videometrics III, Proc. SPIE 2350, pp. 279-290.
Haeberli,P. and Segal,M., 1993. Texture mapping as a
fundamental drawing primitive. In: Proc. of the Fourth
Eurographics Workshop on Rendering, June, Paris, France.
Hedberg, S.R., 1996. Virtual reality at Boeing: Pushing
the envelope. Virtual Reality Special Report, 3(1), Jan,-
Feb., pp. 51-55.
Phillips-Mahoney, D., 1995. Modeling for Virtual
Reality. Computer Graphics World, 18(10) October, pp. 45-
50.
Polis,M.F., Gifford,S.J., and McKeown Jr.,D.M., 1995.
Automating the construction of large-scale virtual worlds.
IEEE Computer, 28(7) July, pp. 57-65.
Rioux,M., 1984. Laser range finder based on
synchronized scanners. Applied Optics, 23, pp. 3837-3844.
Soucy, M., Godin, G., Baribeau, R., Blais, F., Rioux,
M., 1996. Sensors and algorithms for the construction of
digital 3-D colour models of real objects. To appear in:
Proc. IEEE Int Conf on Image Processing, Lausanne,
Switzerland, September 16-19.
Sturman,D.J. and Zelter,D., 1994. A survey of glove-
based input. IEEE Computer Graphics & Applications, 14(1)
January, pp. 30-39.
Stevens, L., 1994. Virtual Reality Now. MIS Press, N.Y.,
N.Y., USA.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996
KEY WOF
ABSTRA!
This paper
structural
A Fujix DS
of a struct
For this c
adjustmer
The result
KURZFA:
In diesen
herkómml
Die Digitall
Flachentr:
Zu Vergl
Selbstkalil
Among all the
close-range pl
contact, and
Basically, de
application o
coordinate me
more epochs
Analytical clo:
level of accure
precision wo
become a re
advantages al
automated de
The developn
reached a hig|
a variety of diff
applications i
test model.
In this pape
compared to 1
