2111 An out-of-core mesh generation method from noisy large point-clouds
Kenji Murakami, Satoshi Fujii, Hiroshi Masuda, Ichiro Tanaka · Sekkei Kougaku, Shisutemu Bumon Kouenkai kouen rombunshuu/Sekkei Kogaku, Shisutemu Bumon Koenkai koen ronbunshu · 2007
Recently, 3D scanners are significantly improved and a state-of-art 3D scanner can measure tens of millions of points in only one scanning. However, most PCs cannot handle such a large number of points because of the limitation of memory space. We intend to build 3D models of a manufacturing facility. Although it is difficult to process the huge data of an entire facility in limited memory space, most components in the facility can be processed if they are separately stored. Therefore, we subdivide a huge point-cloud into sub-sets, generate mesh models from each set of points, and merge mesh models. We applied our technique to point-clouds of a nuclear power plant.