X3D-Earth: Full Globe Coverage Utilizing Multiple Dataset

摘要

U.S. Armed Forces are typically utilizing and paying for commercial proprietary licensed products to visualize geospatial scenes, which nevertheless are themselves derived from government-produced data. This is unsatisfactory. This thesis has developed an open-source, royalty-free method for generating full-coverage 3D globes using the Extensible 3D (X3D) Graphics International Standard. Specifically, this thesis designs and generates robust globe models developing an instance of full global coverage utilizing X3D- Earth. In order to show interoperability and 'Mash up' capabilities, multiple formats are used, including DTED Level 0 and NGA-produced satellite imagery. Imagery and corresponding terrain datasets are preprocessed using image processing and terrain parsing software, creating the X3D-Earth quad-tree tiles into multiple level-of-detail (LOD) file archives. Finally, these pyramidal, locale scenes are grouped and connected to form an overall X3D-Earth globe. Preprocessing, processing and data storage are performed with the NPS Hamming Supercomputer. The result of this work is a methodology for generating X3D- Earth locales that is suitable for massive replication, optimization and reuse. Current results are promising and further work is warranted. The ultimate product is expected to further enable new tactical capabilities, provide direct end-user control of visualization-data pedigrees, and enable improved operational and navigational situational awareness for all deployed warfighters.