ind accuracy 
arison results 
oduced point 
ice (Tables 6, 
jpancies with 
re probably 
arison results 
jarison results 
relying on the 
n interpolated 
surfaces. Such 
but can have 
for complex 
it the. gearbox 
resulting in an 
| the reference 
is a noticeable 
ich cannot be 
ay be linked to 
the efficiency, 
se automated 
es using multi 
hod has been 
known quality. 
white light and 
cases a dense 
with very few 
ent. 
The quality evaluation has shown a significant difference 
between the precision estimates and the accuracy data. It is not 
possible to attribute this to any one cause because of the 
unquantified influence of the surface roughness between the 
different techniques and the method of surface comparison 
used. However, comparison to the CMM reference surfaces has 
demonstrated an agreement to the order of 0.3 mm for all 
datasets. Assessments of variations in the results are also made 
difficult by the fact that the point cloud coverage is different 
between datasets. 
Surface information was not successfully recovered in areas of 
shadow and occlusion caused by the depth discontinuities at the 
particular angle and distance of the images with respect to the 
object surface. Such occlusion problems can be resolved by 
acquiring further images from different locations and 
introducing them in the densification process. Furthermore, it is 
also possible to vary camera exposure and projected image 
intensity to obtain better image quality in both shadow and 
highlight areas. These two additions to the process would 
significantly increase the degree of surface coverage whilst 
maintaining measurement accuracy. 
It is worth noting that a greater number of points can be derived 
from the densification process if a higher image magnification 
is achieved whilst maintaining angle of view. This can be 
achieved with a higher resolution digital camera, such as the 
DCS460 (Papadaki et al, 2001b). 
REFERENCES 
Beraldin J. A., Blais F., Boulanger P., Cournoyer., L., Domey 
J., El-Hakim S.F., Godin G., Rioux M.and Taylor J., 2000, Real 
world modeling through high resolution digital 3D imaging of 
objects and structures, ISPRS Journal of Photogrammetry and 
Remote Sensing, 55 (4), pp. 230-250. 
Boeseman I., Schneider C.T., 2001, On-line 3D measurement 
using inverse photogrammetry, SPIE, 4309, 288-293. 
Clarke, T.A., 1994, An analysis of the properties of targets uses 
in digital close range photogrammetric measurement, 
Videometrics III. Boston, SPIE Vol. 2350, pp. 251- 262. 
D'Apuzzo, N., 2002, Modelling human faces with multi-image 
photogrammetry, Corner, B.D., Pargas, R., Nurre, J.H. (Eds.), 
Three-Dimensional Image Capture and Applications V, Proc. of 
SPIE, Vol. 4661, San Jose, USA, pp. 191-197. 
Deriche R., 1993. Recursively implementing the Gaussian and 
its derivatives. Technical Report 1893, INRIA. Unité de 
Recherche Sophia-Antipolis. 
Forstner W., Gulch E., 1987. A Fast Operator for Detection and 
Precise Location of Distinct Points, Corners, and Centres of 
circular Features. Proceedings of Intercommission Conference 
of ISPRS on Fast Processing of Photogrammetric Data. 
Interlaken, pp. 281-305. 
Gruen, E. Baltsavias, 1988, Geometrically constrained 
multiphoto matching, Proceedings Intercommission Conference 
on Fast Photogrammetric Processing, Interlaken, June 2-4, 
1987. Photogrammetric Engineering and Remote Sensing, Vol. 
54, No. 5, May, pp. 633-641. 
Papadaki H., Robson S, Woodhouse N.W., Chapman D., 2001a. 
Towards the automated extraction of accurate 3D models from 
close range photogrammetric networks, SPIE Electronic 
Imaging, Videometrics, San Jose. 
Papadaki H., Robson S, Woodhouse N.W., Chapman D., 2001b. 
Obtaining accurate dense engineering data sets using an 
integrated close range photogrammetry and machine vision 
solution, Vienna, 5th conference on optical 3D measurement 
techniques, Proceedings. 
Robson S., Shortis M., Ray S., 1999, Vision metrology with 
super wide angle and fish-eye optics, SPIE Videometrics VI, 
San Jose California, Vol. 3641, eds. A. Gruen, S. el-Hakim. 
199-206. 
Siebert J. P., Marshall S. J., 2000, Human body 3D imaging by 
speckle texture projection photogrammetry, Sensor Review 20 
(3), p.p. 218-226. 
VMS: Vision Metrology System, developed by S. Robson, UCL 
and M. Shortis, University of Melbourne. 
Woodhouse N.G., 2000, Geometric models appropriate for 
engineering analysis from vision metrology data, PAD thesis, 
Department of Geomatic Engineering, UCL. 
[1].http://www.optimumvision.co.uk/kodak.htm (accessed 23 
May 2002) 
[2].http://www.npl.co.uk/ (accessed 30 May 2002) 
[3].http://Www.leica-geosystems.com/ims/product/axyz.htm 
(accessed 05 June 2002) 
AKNOWLEDGEMENTS 
The author would like to thank Dr. Stuart Robson, UCL, for his 
assistance in the development of this research as well as the 
painstaking tasks of image acquisition and data processing. 
The author would also like to thank Simon Oldfield, NPL, for 
kindly providing the CMM data. 
—73—