FROM POINT SAMPLES TO SURFACES - ON MESHING AND ALTERNATIVES (original) (raw)

Abstract

Terrestrial laser scanners deliver a dense point-wise sampling of an object's surface. For many applications a surface-like reconstruction is required. The most typical example is the visualization of the scanned data. Traditional approaches use meshing algorithms to reconstruct and triangulate the surface represented by the points. Especially in cultural heritage, where complex objects with delicate structures are recorded in highly detailed scans, this process is not without problems. Often long and tedious manual clean-up procedures are required to achieve satisfactory results. After summarizing our experience with current meshing technology we therefore explore alternative approaches for surface reconstruction. An alternative approach presented within this paper is point splatting. We have developed an algorithm to compute a suitable surfel representation directly from the raw laser scanner data. This results in a speedy and fully automated procedure for surface reconstruction. The properties of the different approaches for surface reconstruction are discussed considering a practical example from the field of cultural heritage. The Panagia Kera in Kritsa near Agios Nikolaos on the island of Crete was chosen as a suitable example.

FAQs

sparkles

AI

What challenges arise from using terrestrial laser scanning for cultural heritage documentation?add

The process demonstrates that inhomogeneous sampling and missing adjacency information often create artifacts in the meshing results, despite achieving smooth surfaces with moderate manual intervention.

How does point splatting compare to traditional meshing methods in surface reconstruction?add

Point splatting avoids issues like non-uniform sampling and adjacency loss, yielding quicker visualization results without heavy manual interaction, albeit at a compromise in surface detail.

What are the primary strategies for deriving triangle meshes from point samples?add

There are two strategies: one computes meshes from unorganized points, while the other utilizes scanning device properties for integration, as seen in Geomagic and Polyworks.

What does the efficiency of the proposed point splatting algorithm reveal about processing times?add

The algorithm processes over two million points in about 100 seconds, achieving a speed of 20,000 points per second, showcasing its efficiency for large datasets.

How does manual point removal affect the quality of the registered point cloud?add

In the case study, manual editing reduced the initial point cloud from 2.8 million to 1.5 million points, indicating significant improvements in data quality for subsequent processes.

Loading...

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

References (10)

  1. Böhm, J., 2005. Terrestrial laser scanning -a supplementary ap- proach for 3d documentation and animation. In: Fritsch (ed.), Photogrammetric Week '05, Wichmann, Heidelberg, pp. 263- 271.
  2. Edelsbrunner, H., 2003. Surface reconstruction by wrapping fi- nite point sets in space. In: J. P. B. Aronov, S. Basu and M. Sharir (eds), Discrete and Computational Geometry -The Goodman-Pollack Festschrift, Springer-Verlag, pp. 379-404.
  3. Kobbelt, L. and Botsch, M., 2004. A survey of point-based tech- niques in computer graphics. Computers and Graphics 28(6), pp. 801-814.
  4. Owen, S. J., 1998. A survey of unstructured mesh genera- tion technology. In: Proceedings 7th International Meshing Roundtable, Dearborn, MI.
  5. Pfister, H., Zwicker, M., van Baar, J. and Gross, M., 2000. Sur- fels: Surface elements as rendering primitives. In: K. Ake- ley (ed.), Siggraph 2000, Computer Graphics Proceedings, ACM Press / ACM SIGGRAPH / Addison Wesley Longman, pp. 335-342.
  6. Rusinkiewicz, S. and Levoy, M., 2000. QSplat: A multiresolution point rendering system for large meshes. In: K. Akeley (ed.), Siggraph 2000, Computer Graphics Proceedings, pp. 343-352.
  7. Soucy, M., 1997. Inovmetric's multiresolution modeling algo- rithms. In: SIGGRAPH '97 Course Notes Multiresolution Sur- face Modeling, Los Angeles.
  8. Soucy, M. and Laurendeau, D., 1995. A general surface approach to the integration of a set of range views. IEEE Transactions on Pattern Analysis and Machine Intelligence 17(4), pp. 344-358.
  9. Weyrich, T., Pauly, M., Keiser, R., Heinzle, S., Scandella, S. and Gross, M., 2004. Post-processing of scanned 3d surface data. In: S. R. M. Alexa (ed.), Eurographics Symposium on Point- Based Graphics.
  10. Zwicker, M., Pauly, M., Knoll, O. and Gross, M., 2002. Pointshop 3d: An interactive system for point-based surface editing. In: SIGGRAPH '02, pp. 322-329.