We proposed an algorithm for the triangulation of 3D cloud points. It was essentially a surface-based approach, but with several new concepts to overcome the computational and memory problems in the available algorithms. We proposed a complete data structure to record the connectivity information among the triangular elements and the cloud points. A subdivision algorithm was proposed also to assign each cloud point into an appropriate cell so as to reduce the search required. A rule-based triangulation process was proposed for the growing of new triangles. Six types of triangles were created based on the relationship between the best point and the existing elements. The computational speed of the proposed algorithm was quite efficient. Several examples were provided to demonstrate the feasibility of the proposed algorithm for data up to several hundred thousand points.
|Number of pages||12|
|Journal||Journal of the Chinese Society of Mechanical Engineers, Transactions of the Chinese Institute of Engineers, Series C/Chung-Kuo Chi Hsueh Kung Ch'eng Hsuebo Pao|
|State||Published - 2005|
- Reverse engineering
- STL data