Handling huge and complex 3D geometries with Semantic Web technology

Florence Heri-Tech - The Future of Heritage Science and Technologies

Bonsma, Peter^1 ; Bonsma, Iveta^1 ; Ziri, Anna Elisabetta^2 ; Iadanza, Ernesto^3 ; Maietti, Federica^4 ; Medici, Marco^4 ; Ferrari, Federico^4 ; Sebastian, Rizal^5 ; Bruinenberg, Sander^5 ; Lerones, Pedro Martin^6

RDF Ltd., Bulgaria^1

Nemoris Srl., Italy^2

Consorzio Futuro in Ricerca, Italy^3

University of Ferrara, Department of Architecture, Italy^4

DEMO B.V., Netherlands^5

Fundación CARTIF, Spain^6

关键词: Collaborative research projects;    Complex 3D geometry;    Cross-disciplinary research;    Cultural heritage objects;    Cultural heritages;    Geometric information;    Performance issues;    Semantic Web technology;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/364/1/012041/pdf
DOI  :  10.1088/1757-899X/364/1/012041

来源: IOP

【 摘 要 】

In INCEPTION, a European collaborative research project, a Heritage BIM (H-BIM) ontology is being developed to store all relevant semantic data concerning cultural heritage objects. Similar to other projects dealing with storing semantics, one of the major questions is whether, and if yes, how should geometry be stored using semantic web technology. The INCEPTION cross-disciplinary research consortium chose to allow the storage of all relevant geometric information using semantic web technology. The alternative is to store geometry in a different way, or storing only the aggregated parts of geometry, for example through bounding box representations. The geometry is originally generated by a CAD/BIM system and, as we are dealing with Cultural Heritage, in many cases it is derived from 3D point clouds. These result in a large amount of 3D data to be stored using semantic web technology. A well-known issue is that the performance of databases and inferencing engines for semantic web data drops considerably when the data grows to very large sizes. This paper explains how the performance issues on these large sets of geometric data can be solved while still being able to use the databases, inferencing engines, and the geometric data effectively.