(c) ; (d) 
Fig. 5: Modelling an industrial site (a) Point Cloud (b-c) Images 
(d) Final Model 
much more general, as it uses both point cloud and image 
measurements in big adjustment. 
This integrated adjustment was applied to the test scenario 
shown in Fig. 5. Only cylinders, boxes and tori were used from 
the catalogue of CSG objects. The results of the fitting are 
shown as a 3D Model in Fig. 5(d). 
5. CONCLUSIONS 
We have presented a modelling technique for fitting CAD 
models described as CSG primitives to measurements in images 
and point clouds. While the point clouds are excellent for 
automatic object recognition, the comparison of improvement in 
the standard deviation of the estimated parameters clearly shows 
that images have a complementary role as they provide more 
information on the edges and help fix the bounds of models 
where point clouds fail to do so. In future we plan to extend the 
fitting procedure from point measurements to edge and curve 
measurements in images. Similarly different strategies for 
automatic object recognition in industrial environments using a 
combination of imagery and point clouds will also be 
investigated. 
REFERENCES: 
ACIS, 2004. The 3D ACIS® Modeler (ACIS) www.spatial.com 
(Last accessed April 29, 2004) 
Besl, P., and McKay, N., 1992. A method for Registration of 
3D shapes. In Pattern Analysis and Machine Intelligence, 14(2). 
239-256. 
Bjork, A., 1996. Numerical Methods for Least Squares 
Problems. Society for Industrial and Applied Mathematics, 
Philadelphia. 
Dennis, J.E. and Schnabel, R.B.. 1996 Numerical Methods for 
Unconstrained Optimization and Nonlinear Equations, SIAM, 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004 
Second Ed., Philadelphia. 
Dijkman, S.T., Heuvel, F.A. van den, 2002. Semi automatic 
registration of laser scanner data. International Archives of 
Photogrammetry and Remote Sensing, 34(5), pp 12 - 17 
Ermes, J.P.A.M., Heuvel, F.A. van den, Vosselman, G., 1999. 
A Photogrammetrie Measurement Method using CSG Models. 
International Conference proceeding on Measurement, Object 
modelling and documentation in architecture and industry, 
Thessaloniki, 7-7-1999, IAPRS, 32(5), pp. 36-42. 
Golub, G.H., Van Loan, C.F., 1996. Matrix Computations, 3rd 
Ed. The Johns Hopkins University Press. 
Hoover, A., Jean-Baptiste, G., Jiang, X., Flynn, P.J., Bunke, H., 
Goldgof, D.B., Bowyer, K., Eggert, D.W., Fitzgibbon, A., and 
Fisher, R.B., 1996. An experimental comparison of range image 
segmentation algorithms. IEEE Transactions on Pattern 
Analysis and Machine Intelligence, 18 (7), 673-689. 
Laser Scanner Survey, 2003 (Conducted by POB Magazine). 
http://Www.pobonline.com/FILES/HTML/PDF/01033d-laser- 
survey.pdf (Accessed March 2, 2004). 
Lowe, D. G., 1991. Fitting parametrized three-dimensional 
models to images. PAMI, 13(5), 441-450. 
Min, A, Powell, M.W., Bowyer, K.W., 2000. Progress in 
automated evaluation of curved-surface range image 
segmentation. In International Conference on Pattern 
Recognition (ICPR), Barcelona, Spain, pp.644-647. 
Nourse, B., Hakala, D., Hillyard, R., Malraison,P., 1980. 
Natural quadrics in mechanical design.In Proc. of Autofact 
West, Anaheim, CA., 363-378. 
Petitjean, S., 2002. A survey of methods for recovering quadrics 
in triangle meshes. In ACM Computing Surveys, 34(2). pp. 
211-262. 
Press, W.H., Teukolsky S.A., Vetterling W.T., Flannery B.P., 
1996. Numerical Recipes in C: The Art of Scientific 
Computing, 2nd ed. Cambridge University Press, Cambridge, 
U.K. 
Rabbani, T., Heuvel, F.A. van den, 2004. Methods For Fitting 
CSG Models to Point Clouds and Their Comparison. The 7th 
IASTED International Conference on Computer Graphics And 
Imaging, August 17-19; 2004, Kauai, Hawaii, USA. 
Requicha, A., Voelcker, H., 1982. Solid modeling: a historical 
summary and contemporary assessment. IEEE Comput. Graph. 
Appl. 2(2), 9-24. 
Shoemake, Ken, 1985. Animating rotation with quaternion 
curves. Proceedings of the 12th annual conference on Computer 
STAR-Services and Training through Augmented Reality. 
http://www.realviz.com/STAR/objectives.htm (Accessed Nov 
11, 2003) 
Vosselman, G., Tangelder, J. W.H., Ermes, P., Heuvel van den, 
F. A, 2003. CAD-Based Photogrammetry for Reverse 
Engineering of Industrial Installation. Computer-Aided Civil 
and Infrastructure Engineering, Vol. 18 pp. 264-274 
  
A PE 
* Dept 
h Inst. 
KEY W( 
ABSTR/ 
The res 
DEM is 
problem 
to find 
photogr: 
projecte 
and proj 
mm in h 
Accurate 
boned px 
safety foi 
potentiall 
processin 
Current 2 
false-det« 
detector : 
ray absor 
patterns 
between 
Although 
be used, 
detecting 
alone. 
An appro 
use an 2 
thickness 
recogniti: 
2.f 
The prese 
generatin 
an imagi 
chicken f 
(DEM). 1 
].