3D VIEWPOINT OPTIMIZATION OF TOPOLOGICAL RELATIONSHIPS: APPLICATION TO 3D CADASTRE FOR VISUAL EASEMENT VALIDATION

Abstract. Offering optimum 3D viewpoint to user can be attractive in relieving occlusion in 3D scene. This could be much relevant for the visualization of 3D cadastral systems since they constitute complex datasets including both physical and legal objects while users are operating a number of visual tasks that require precise outlook. However, 3D viewpoint usability has yet to be evaluated to demonstrate its relevance in accomplishing given end user’s visual tasks. Hence, in this research project, the focus is set on visual identification of 3D topological relationships (disjoint and overlap) as it is one of the main users’ requirements in 3D cadastre. To this end, this paper addresses this issue using a virtual 3D model of the Planetarium Rio Tinto Alcan (Montreal city) in which property issues take place, especially regarding the easement validation procedure. Empirical tests have then been administrated in the form of interviews using an online questionnaire with university students who will specifically address such issues in their professional career. The results show that a 3D viewpoint that maximizes 3D disjoined or overlapped geometric objects’ view area within the viewport significantly outperforms traditional combined software points of view in visually identifying 3D topological relationship. This paper also suggests that user’s inexperience in 3D cadastre reduces visual task efficiency when visually identifying 3D topological relationship among overlapped geometric objects. Eventually, this study opens up new perspectives on 3D topological relationships modeling and visualization.