ORGANIZING SMART CITY DATA BASED ON 3D POINT CLOUD IN UNSTRUCTURED DATABASE – AN OVERVIEW
Keywords: 3D Point Cloud, Big Data, Database Management, Data Model, Unstructured Database
Abstract. The concept of the 3D smart city is an integration of smart cities and information technology. One of the data sources of a smart city is point cloud data that are produced from various data acquisition tools such as LiDAR, Terrestrial Laser Scanning, and Unmanned Aerial Vehicle. Due to the large size of point cloud data input, traditional databases could not handle the data efficiently. Alternatively, unstructured databases have become an option. Furthermore, data for smart city applications are considered being complex and large. Storing data in the unstructured database can easily be retrieved from various front ends such as web and mobile devices. However, unstructured databases do not have fixed schema and data types that often limit the uses of 3D point cloud data in relational databases. There are four categories of the data model in the unstructured database: document store, key-value, column store, and graph store. Each of the categories has different characteristics and approaches to handling data. Thus, this paper aims to summarise an overview of each category and determine the most suitable data organisation and environment for a 3D point cloud of a smart city. The overview will aid the developer or user select and comparing available data models in the unstructured database to handle 3D point clouds.