fig congress 2018 - 3D Cadastres Best Practices, Chapter 4: 3D Spatial DBMS for 3D Cadastres
Karel Janečka, Sudarshan Karki, Peter van Oosterom, Sisi Zlatanova, Mohsen Kalantari
Subdivision of land parcels in the vertical space has made it necessary for cadastral jurisdictions to manage cadastral objects both in 2D as well as 3D. Modern sensor and hardware capabilities for capture and utilisation of large point clouds is one of the major drivers to consider Spatial Database Management Systems (SDBMS) in 3D and organisations are still progressing towards it. 3D data models and their topological relationships are two of the important parts of 3D spatial data management. 3D spatial systems should enable data models that handle a large variety of 3D objects, perform automated data quality checks, search and analysis, rapid data dissemination, 3D rendering and visualisation with close linkages to standards. This chapter asserts that while there has been work done in defining 2D and 3D vector geometry in standards, it is still not sufficient for 3D cadastre purposes as 3D cadastral objects have a much more rigorous definition. The Land Administration Domain Model (LADM), which is an ISO Standard, addresses many of the issues in 3D representation and storage of 3D data in a database management system (DBMS). The chapter further discusses the various approaches to storing 3D data such as through voxels, or point cloud data type and elaborates on the characteristics of a 3D DBMS capable of storing 3D data. Approaches for spatial indexing to improve the fast access of data and the various available options for a 3D geographical database system are presented. Several spatial operations on and amongst 3D objects are illustrated with linkages to the current standards including the LADM. Next, construction of 3D topological and geometrical models based on standards and including their characteristics is discussed. Current 3D spatial database managements systems and their characteristics, including some comparison between selected DBMS including the hardware capabilities are elaborated in detail. Finally, the chapter proposes a 3D topology model based on Tetrahedron Network (TEN) synchronised with LADM specifications for 3D cadastral registration. This topological model utilises surveying boundaries to generate 3D cadastral objects with consistent topology and rapid query and management capabilities. The definition for validation of 3D solids also considers the automatic repair of invalid solids. Point cloud and TEN related data structures available in SDBMSs are also investigated to enable storage of non-spatial attributes so that database updates would store all spatial and attribute information directly inside the spatial database.
Only personal, non-commercial use of this document is allowed.