Digital Mapping Laboratory
Carnegie Mellon University
TitleConstruction of Synthetic Environments for Advanced Distributed Simulation[color]
AuthorsDavid M. McKeown, Jr., George Bulwinkle, Steven Douglas Cochran, Wilson Harvey, J. Chris McGlone, Michael Polis, Jefferey A. Shufelt
AbstractThis final report presents an overview of research and advanced development in image understanding and automated cartography, from April 1995 to February 2000, to support the efficient representation and rapid construction of virtual world databases. Topics discussed include advances in cartographic feature processing to allow and improve object-space image representation and geometric constraints; the automatic and semi-automatic extraction of buildings and roads from aerial imagery; hyperspectral image analysis for material classification; the fusion of information extracted from different extraction methods, the use of rigorous metrics and methods to evaluate cartographic feature extraction systems, and the use of aerial and terrestrial imagery to phototexture site models for visualization using photogrammetric principles.The organizing theme for this program of research was to support the construction, intensification and generalization of geospatial databases to support synthetic environments and real-world simulation. As a component of an overall investment in the DARPA Synthetic Theatre of War (STOW) demonstrations, this Synthetic Environments (SE) program provided a fundamental investment in the technology infrastructure necessary to more effectively automate the generation of source geospatial data required for rapid construction of virtual worlds. In addition to a focus on automated cartographic feature extraction, the CMU MAPSLab pursued novel research and development of terrain database representations which lead to the adoption of integrated triangular irregular networks (ITINS). ITINS have become the predominante underlying representation for several STOW exercise databases, as well as influencing support within modular semi-automated forces (ModSAF) compact terrain database representation (CTDB). for the ITIN representation. Through several technical interchanges (TINDays) organized within the RCVW program, ITIN technology was adopted by all major commercial terrain database construction companies as an integral component of their COTS software strategy. In the body of this report we provide an overview of the range of activities sponsored under this contract. This final report provides pointers to detailed technical reports, presentations, and ancilliary information that fully describe the innovative research and development performed at the CMU MAPSLab under this RCVW contract. Keywordsdigital photogrammetry, terrestrial imagery, monocular building extraction, stereo analysis, road network extraction, multispectral image analysis, data fusion, evaluation metrics, database construction, virtual world construction, simulation.
Select Version
|
maps+webmaster@cs.cmu.edu