Geospecific Night Imagery for Real-time Training Simulators

To meet future JSF programmatic needs and to augment SOF mission planning and training capability, SimWright proposes eight trade studies and a risk mitigation program leading to a system design to generate geo-specific terrain databases. The eight trades studies are characterized as: - Night Imagery Sources Study - Street Light Feature Extraction Feasibility Study - Tower Feature Extraction Feasibility Study - Imagery Correlation Feasibility Study - Moving/Blinking/Occlusion Detection Feasibility Study - Shader Feasibility Study - DBGS Input Requirements Study - Day for Night Feasibility Study. Preliminary investigation suggests no known sources of night time geo-specific terrain data. Having no night time geo-specific data is an enormous risk to developing stereographic feature extraction algorithms for geo-specific night time terrain databases. SimWright will have approximately 1 square mile of night imagery of Destin FL., prior to Phase I award. Imagery correlation can begin right at the beginning of the program. SimWright’s CityClick framework is extensible for rapid prototyping of light feature extraction capability. We will investigate both manual and automatic feature extraction for major trade analyses, algorithmic verification, and software prototyping. DBGS formats, input and compatibility issues with IG pipelines will be investigated as part of overall risk reduction program.