ystem was fixed
ance was about
with 24 control
of the reduced
case, the media
. By Comparing
ior orientation
ted results are
nodel is correct.
trol points) are
inal vector RMS
(10 points) are
=16.1 mm.
VETS
ht camera
598.70
280.62
213.68
1.078
00.8951
00.6376
)7.8027
0.0328
0.0084
0.0096
).00001
.00010
0.04
0.02
ol frame merged
he uncontrolled
reduced central
to compute the
in Table 2. The
fferent from the
caused by the
rections and the
parameters are
y tracing model.
1at the RMS of
ind m;-2.1mm,
used for this
odel, interested
Table 2. Real sea site test
Parameter Left camera Right camera
f (pixel) 563.92 569.5]
Xp (pixel) 320.02 317.03
Yp (pixel) 223.34 229.71
Sy 1.075 1.078
X0(m) 101.1556 101.5426
Y0(m) 100.3727 100.3716
Z0(m) -98.1268 -98.3053
o (rads) -0.0690 -0.0453
$ (rads) -0.2890 -0.2151
K (rads) -0.0624 -0.0131
K1(10:2) -0.0001 -0.0001
K2(10°%) 0.00001 0.00001
Pj (1074) -0.006 -0.02
Pa(10-^) -0.020 -0.021
6. CONCLUDING REMARKS
The underwater photogrammetric models for extracting
quantitative spatial information of underwater objects using
CCD stereo images have been researched. In addition, a PC
based UDPS has been developed to perform the underwater
photogrammetric processing. From the testing results, the
following conclusions can be drawn:
e The newly derived reduced central perspective model was
successfully applied to compute the calibration parameters
of the imaging system. This method provides us an
efficient way to address the problems of acquisition of
approximate values, especially in uncontrolled underwater
environmental situations.
e [tis possible to use the 3D optical ray tracing technique to
describe the imaging procedure and to construct a rigorous
photogrammetric model with multiple refraction and
multi-lens of the imaging system.
* The designed system using CCD cameras and mobile
photogrammtric method are efficient for underwater data
acquisition. It is a promising way to perform a non
seafloor control survey.
* The method applied allowed an accuracy of 0.5-1.0 cm
along the x- and y-direction and 1.0-3.0 cm along the z-
direction in the object space.
The further research will focus on the error and reliability
analysis of the imaging system under different underwater
conditions. The integration of multiple sensors for objects
measurement will be conducted. Digital image classification
and pattern recognition for specific objects, e.g., fish species,
Will be carried out.
7. ACKNOWLEDGMENTS
The support from the Canadian Hydrographic Services (CHS),
Department of Fisheries and Oceans (DFO), and the National
Sciences and Engineering Research Council of Canada
(NSERC) is gratefully acknowledged.
8. REFERENCES
Adams, L. P., 1982. Underwater Analytical Photogrammetriy
Using Non-metric Cameras, Int. Archives of Photogrammetry,
25(5):12-22.
Fryer, J. G. and C. S. Fraser, 1986. On the Calibration of
Underwater Cameras, Photogrammetric record, 12(67): 73-85
Kotowski, R., 1988. Phototriangulation in Multi-media
photogrammetry, Int. Archive of ISPRS, Com. V, Part 27: 324-
334.
Leatherdale, J. D. and D. J. Turner, 1991. Operational
experience in underwater photogrammetry, ISPRS J. of P&RS,
46, pp. 104-112.
Lenz, R., 1987. Lens Distortion Corrected CCD-camera
calibration Points for Real-Time 3D Measurements. Proc. of
Inter Commmission on “Fast Processing of Photogrammetric
Data”, Interlaken, pp. 60-67.
Li, H., 1995. Quantitative Analysis of Underwater Stereo
Video Images, M.Sc. Thesis, The University of Calgary.
Rinner, K., 1969. Problems of two medium photogrammetry.
PE, 35(2): 275-282
Tumer, D. J., D. Yule, and J. Zanre, 1992. A Real Time
Photogrammetry System for Underwater and Industrial
Applications. Int. Archives of ISPRS, Part 28(5):507-513.
Welham, L. G., 1984. The Development of an Underwater
Measuring Capability Using Photogrammetric Technique. Int.
Archives of ISPRS, 25(5): 757-766
323
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996
