SplitSFC: A database solution for massive point cloud data management
Y. Cai (TU Delft - Architecture and the Built Environment)
B.M. Meijers – Mentor (TU Delft - Architecture and the Built Environment)
P.J.M. van Oosterom – Graduation committee member (TU Delft - Architecture and the Built Environment)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Point cloud data are gaining more importance in the Geomatics domain, with the development of modern sensing technologies like LiDAR and photogrammetry. The size of massive point clouds are growing, and the performance of traditional database solutions for its data management become insufficient. It remains a huge challenge for researchers to come up with a data management solution that handles the huge volumes of data, while providing standardized functionalities.
Space Filling Curve (SFC) has been explored as a good spaital access techniques for organizing point clouds. Existing SFC-based database solutions include PlainSFC, HistSFC, etc. However, they use a flat table to store point records and it is not compact for massive point clouds. In this thesis, a SFC-based database solution that manages point cloud in blocks is proposed. The purpose is to improve the performance of current point cloud database solutions, especially with storage space. This model organizes the point clouds based on Space Filling Curve, and innovatively splits each SFC key to a head and a tile. The points with the same SFC head are placed in the same block. SFC tails and other property dimensions are stored as arrays in other columns.
Compared with the pgPointCloud and Oracle SDO_PC, the intermediate SplitSFC prototype does not show significant advantage in storage and data retrieval efficiency so far. However, it is fair to believe that with the improvement of algorithms and implementations, it has the potential to be an approximate and efficient point cloud data management solution.