ABSTRACT:Recently, services and systems that deal with indoor spatial information are increasing. Each service or system adopts a data model that can store necessary indoor space data according to its purpose. However, since the content of indoor spatial information that can be expressed by each data model is differ and limited, it is necessary to exchange information between the systems in order to use rich indoor spatial data. OGC has published IndoorGML as the standard for exchange of indoor spatial information data between systems. To use IndoorGML as an exchange format, the software which supports IndoorGML construction is fundamental. But there are several limitations in the previous IndoorGML data editing tools. There is no editing tool that can generate all the features which are defined by IndoorGML. If users want to generate IndoorGML data, they need to consider the requirements of the IndoorGML. In this study, we implemented InFactory, which is a IndoorGML generation tool based on RESTful API supporting users to easily construct IndoorGML data. Users can easily create IndoorGML without knowledge on the schema and requirements of IndoorGML using InFactory. In addition, developers on IndoorGML data construction tools such as GUI editors do not have to implement duplicated IndoorGML generation program for their systems. Using Java API that supports CRUD on IndoorGML data, users can also deal with IndoorGML data in their applications.

ABSTRACT:The 3D triangle mesh is widely used to represent indoor space. One of widely used methods of generating 3D triangle mesh data of indoor space is the construction from the point cloud collected using LIDAR. However, there are many problems in using generated triangle mesh data as a geometric representation of the indoor space. First, the number of triangles forming the triangle mesh is very large, which results in a bottleneck of the performance for storage and management. Second, no consideration on the properties of indoor space has been done by the previous work on mesh simplification for indoor geometric representation. Third, there is no research to construct indoor spatial standard data from triangle mesh data. For resolving these problems, we propose the a method for generating triangular mesh data for indoor geometric representation based in the observations mentioned above. First this method removes unnecessary objects and reduces the number of surfaces from the original fine-grained triangular mesh data using the properties of indoor space. Second, it also produces indoor geometric data in IndoorGML -an OGC standard for indoor spatial data model. In experimental studies, we present a case study of indoor triangle mesh data from real world and compare results with raw data.