Automated Feature Extraction Capabilities for the Development of High-Resolution GEOINT Feature Data and Constructing Correlated Databases

The Phase I program demonstrated the feasibility of using optical and LIDAR data to accurately extract structures, vertical obstructions, power lines, and road networks. In the Phase II program, TSC proposes to further develop the extraction techniques, and to develop prototype software at TRL level 6. The software will use a knowledge-based expert system (KBES) for classification, trained with initial user interaction. The KBES methodology facilitates the fusion of multiple sensors, and also adapts easily to different combinations of sensor coverage. Features will be extracted in a vector format consistent with SOCOM’s Common Database (CDB) and will include structures, vertical obstructions, power line networks, road networks, and waterway networks. The software will be designed to operate in SOCOM’s Database Generation Facility (DBGF), and will interoperate with the other DBGF tools in the database generation process.