Towards 3D Metadata for Discovering 3D Geospatial Models

Abstract

Accessing 3D geospatial models, eventually at no cost and technical restriction, is an important issue among the user communities. However, many of the data infrastructures which try to provide access to 3D geospatial data have been originally initiated for publishing 2D resources. To estimate the ability of geo-portals for dissemination of 3D data, metadata, which is usually the foundation of data discovery, is a key component. An inventory on eight geo-portals and commercial websites shows that the metadata which refers to 3D information is very different from one portal to another and among the similar 3D resources in the same portal. The inventory covers more than 900 datasets whose metadata is affiliated with elevation. Three of the inventoried portals do not consider 3D model by any definition. The rest refer to 3D models by different and proprietary concepts and definitions causing confusion and misinterpretation. Motivated to help users with discovering the appropriate 3D geospatial models they need, this research project questions the shortcomings of the current metadata. Correspondingly, it offers to improve the integrity of explicit 3D information in a proposed 3D metadata. The main structure of the proposal comprises 21 genuine metadata classes categorized with regard to the accessibility (contextual and structural) and availability aspects of 3D geospatial data. The proposed 3D metadata is compared with the metadata schema of the Discovery Portal of the Canadian Geospatial Data Infrastructure (CGDI) which partially conforms to North American Profile (NAP) of ISO-19115. The comparison analyzes both metadata schemas against three simulated scenarios about discovering specific types of 3D geospatial datasets. The preliminary results show that our proposal constitutes six new 3D metadata concepts as well as adjusting several others regarding 3D geospatial data.