Inference-based creation of synthetic 3D content with ontologies

Abstract

Creation of synthetic 3D content is typically a complex task, covering different geometrical, structural, spatial and presentational elements. The available approaches enable 3D content creation by programming or visual modeling of its elements. This demands expertise in computer science from content authors, limits the possibilities of using domain knowledge, and requires to determine all content details within the content creation process. In this paper, we propose a new method of 3D content creation, which is based on ontologies. Ontologies are the foundation of the semantic web, enabling the use of knowledge in various domains. In the proposed method, the use of ontologies facilitates 3D content creation by domain experts without skills in programming and computer graphics. In addition, due to the use of ontologies, the method enables automated reasoning, liberating the authors from determining many elements of the created content, which can be inferred by the content generation algorithm on the basis of explicitly specified content elements. The method has been implemented and evaluated. It simplifies 3D content creation in comparison to the available approaches by reducing the number of activities that must be completed by the content authors. Hence, the proposed method can increase the use of 3D content in different application domains.