Automating the Production of Terrain Databases Navy SBIR FY2006.2 Sol No.: Navy SBIR FY2006.2 Topic No.: N06-144 Topic Title: Automating the Production of Terrain Databases Proposal No.: N062-144-0139 Firm: Sentinel AVE LLC 531 Main Street No. 1111 El Segundo, California 90245-3060 Contact: Ulrich Neumann Phone: (310) 704-3579 Web Site: http://www.sentinelave.com Abstract: This proposal builds upon many years of research and experience in the problems of sensor fusion, and 3D modeling. Powerful modeling and data fusion systems, previously developed by the investigators at USC and at Sentinel, will be extended with new capabilities to produce a step change in the automation of modeling and feature extraction for geospatial database systems. The new algorithms are based on our recent explorations into perceptual grouping and classification with sensor cues, feature invariants, and machine learning. Specifically, we integrate the mathematics of tensor voting, Gabor wavelets, and feature-based recognition to automate the extraction of geometric models and terrain features from aerial images and Lidar data. Early results show the feasibility and advantages of our approach. The integration of these methods to attack the feature extraction and modeling problems has never been attempted, to the best of our knowledge; therefore our effort brings a new approach to the geospatial database problem, with the potential for making significant improvements in automation, robustness, and throughput. Benefits: Success in this research will have a significant impact on the development of geospatial databases from sensor data. The resulting system will enable automatic (or minimal supervision) extraction of 3D models and terrain features from aerial images and Lidar. These models and features are used in databases for a variety of military, law enforcement, and civilian applications. Military services require rapid (even real-time) creation of 3D models and feature extraction to support change detection, mission planning, training, and real-time situational awareness. Increased automation will enable a simplified and faster work flow that allows non-experts and non-artists to rapidly fuse sensor data to produce accurate and consistent terrain models and databases in shorter time than possible with current methods. Our results will also have high commercial impact as recent Internet products from Google and Microsoft (e.g., Google Earth) have stimulated a need for the rapid production and maintenance of 3D urban geospatial data. In the commercial GIS, navigation, and Internet mapping services, there is a need for better tools to create and maintain 3D models and databases as a substrate for archived and real-time data to aid planning, training, and decision-support. In the government sector, city planners, architects, and emergency responders would benefit from widely available 3D models to extend the utility of existing 2D GIS systems to the third-dimension (height) to support planning and analysis in urban areas. Academic geographers, civil engineers, and public policy scholars could use models of urban areas for simulations and analysis. Lastly, the US film and computer games industries face increasing cost competition from overseas production of scene and character models. Modeling now dominates production costs in the games industry, overtaking programming cost in recent years, and projected to grow further as scene realism requirements approach those of film production. Return