PlantXML: A DATA EXCHANGE FORMAT FOR TREE INFORMATION

Abstract

Urban tree inventories are typically managed in a 2D GIS (Geographical Information System) database, whereby each tree is represented by a 2D point location, along with its respective 1D semantics such as species information, girth and height measurements, and regular monitoring information of maintenance and health. Other spatial information, particularly tree branching structure and growth space, can be extracted from LiDAR (light detection and ranging) point cloud data by extensive computational processing. However, this extracted information is not technically possible to store in the 2D GIS database which typically has no avenue for storing topological or hierarchical data within a single entity. We propose PlantXML, a novel lightweight eXtensible Markup Language (XML) file format to store the semantics and essential LiDAR-derived spatial information of an individual tree. In this paper, we demonstrate the possibility to extract and store the tree data into the PlantXML format, as well as the reconstruction an individual trees to a certain level of granularity from the PlantXML file. More importantly, over the time, the collection of PlantXML files can be easily managed in a database and processed by machines for further analysis, automated tree growth tracking and other tree management protocols.