Claus SCHEIBLAUER / Michael WIMMER
(TU Wien, Wien, Austria)

Abstract: In recent years the use of remote sensing devices like laser scanners in the documentation of excavation sites or cultural heritage sites has led to huge point cloud models from these sites. These data sets may cover complete sites including galleries, corridors, halls, and open places. Orienting oneself in the point cloud becomes a challenge, if one is not familiar with the layout of the site. Therefore we propose a graph-based guidance system to show tourists round the point cloud models. The tourists can navigate interactively through the point cloud, but they are tied to a predefined 3D graph which represents the possible ways, and which connects the points of interest.
Innovations: A graph-based guidance system in huge point clouds. The point clouds are not pre-rendered, but rendered online. The user can look around freely, but the navigation path is tied to a predefined 3D graph representing the possible ways.
Methodology/Approach: In a preprocessing step a 3D graph is created manually. It consists of nodes (junctions) and edges (ways). A node can have an arbitrary number of edges connected to it. When a user moves along an edge and he arrives at a node, navigation stops and he can choose which edge to follow. All rendering of points is done in real-time with OpenGL. The graph can be visualized with semi-transparent cylinders and spheres for the edges and nodes, respectively.
Results: We created a 3D graph with about 180 nodes inside the Domitilla point cloud model (scanned by the FWF funded START project “The Domitilla Catacomb in Rome. Archaeology, Architecture and Art History of a Late Roman Cemetery”). The graph covers about 15 percent of the 12 kilometers of galleries in the catacomb. Currently we are working to include all points of interest into the graph.

Keywords: point clouds, user guidance, interactive