Figure 15. Model of the room top view 
8. CONCLUSION 
This work was realised to improve a room model despite of 
limited access. Except the specially designed arm carrier, we 
used on-the shelf equipment (camera, light, software) for the 
modelling. 
We obtained an accurate enough 3D model of the room to start 
the decommissioning mission. It can be used to select an 
obstacle free path to develop robotics means towards the tank to 
be inspected and emptied. 
Pyramide capacities to specify and use structure and 
geometrical hypotheses within the modelling process brought an 
efficient compensation for constrained viewpoints. Pyramide 
showed also a high flexibility to adapt the modelling strategy in 
order to cope with erroneous hypotheses and unexpected 
situations, such as obstacles in the way, which altered camera 
position estimation. Many hypotheses were corrected on the 
initial model elaborated formerly. 
The Pyramide model can easily be completed with gamma 
metric information. Furthermore it can be used straight to 
supervise telerobotics missions in the room thanks to its 
connection to robotics-CAD software. 
9. BIBLIOGRAPHY 
Bürger, T. and Busch, W., 2000. Using knowledge about shape 
and position of plant elements in photogrammetric as-built 
documentation. JAPRS, Amsterdam, The Netherlands, Vol. 33, 
Part B5, pp. 107-113. 
Debevec, P., Taylor, C. and Malik, J., 1996. Modeling and 
rendering architecture from photographs: a hybrid geometry- 
and image-based approach. In: SIGGRAPH ’96 Conference 
Proceedings. pp. 11-20. 
Ermes, P. 2000. Constraints in CAD models for reverse 
engineering using photogrammetry. IAPRS, Amsterdam, The 
Netherlands, Vol. 33, Part B5, pp. 209-215. 
Even P. and Marcé, L., 1988. Manned geometric modelling for 
Computer Aided Teleoperation. In: International Symposium on 
Teleoperation and Control. The Ergonomics Society, Bristol, 
UK, pp. 113-122. 
Even, P., Fournier, R. and Gelin, R., 2000. Using structural 
knowledge for interactive 3-D modeling of piping 
environments. In: International Conference on Robotics and 
Automation. IEEE, San Francisco, California, pp. 2013-2018. 
Even, P. and Malavaud, A. 2000. Semi-automated edge 
segment specification for an interactive modelling system of 
robot environments. JAPRS, Amsterdam, The Netherlands, Vol. 
33, Part B5, pp. 222-229. 
Hrabácek, J. and van den Heuvel, F., 2000. Weighted geometric 
object constraints integrated in a line-photogrammetric bundle 
adjustment. /APRS, Amsterdam, The Netherlands, Vol. 33, Part 
B5, pp. 380-387. 
Koehl, M. and Grussenmeyer, P., 2000. Semantic aspects for 
3D modeling. IAPRS, Amsterdam, The Netherlands, Vol. 33, 
Part B4, pp. 507-514. 
Liedtke, C.E., Grau, O. and Growe, S., 1995. Use of explicit 
knowledge for the reconstruction of 3-D object geometry. In: g^ 
Int. Conf. On Computer Analysis of Images and Patterns. 
Prague, Czech Republic, pp. 580-587. 
McGlone, C., 1995. Bundle adjustment with object space 
geometric constraints for site modeling. In: Integrating 
Photogrammetric Techniques with Scene Analysis and Machine 
Vision II. SPIE, Vol. 2486, Orlando, Florida, pp. 19-21. 
Poulin, P., Ouimet, M. and Frasson, M.C., 1998. Interactively 
modeling with photogrammetry. In: Proceedings of 
Eurographics Workshop on Rendering Techniques. Vienna, 
Austria, pp. 93-104. 
van den Heuvel, F., 1998. 3D reconstruction from a single 
image using geometric constraints. ISPRS Journal of 
Photogrammetry and Remote Sensing. Vol. 33 (6), pp. 354-368. 
—180- 
em —i 
So Mo oM m8 0765 (C09 tM se 
pet + 
de ^ D 
Md o -—D  — AA