Arnaud BLETTERER | Frédéric PAYAN | Marc ANTONINI | Anis MEFTAH
Nowadays, LiDAR scanners are able to digitize very wide historical sites, leading to point clouds composed of billions of points. These point clouds are able to describe very small objects or elements disseminated in these sites, but also exhibit numerous defects in terms of sampling quality. Moreover, they sometimes contain too many samples to be processed as they are. In this paper, we propose a local graph-based structure to deal with the set of LiDAR acquisitions of a digitization campaign. Each acquisition is considered as a graph representing the local behavior of the captured surface. Those local graphs are then connected together to obtain a single and global representation of the original scene. This structure is particularly suitable for resampling gigantic points clouds. We show how we can reduce the number of points drastically while preserving the visual quality of large and complex sites, whatever the number of acquisitions.