An Innovative Accurate Predictive 3D Terrain Modeling

The U.S. Air Force is looking for innovative technologies providing an autonomous helicopter sensor-display system for brown-out landing conditions. In response AGNC proposes to develop and document an innovative Accurate Predictive 3D Terrain Modeling (AP-3DTM) system and software toolsets. The AP-3DTM is based on: 1) on-the-fly 3D model generation deploying stereo-cameras with accuracy estimation and quality assurance; 2) modeling the helicopter position geospatial refinement based on AGNC’s coremicro AHRS/GPS/INS Integration Unit sensors measurements; 3) AGNC’s robust Kalman-Filter based helicopter position prediction; 3) optimal model visualization on 3D autostereoscopic or pilot-helmet display. Hazards detection is a concern for which AGNC provides an automated pattern recognition solution. The slope of the landing terrain is also computed to aid the pilot in the execution of a safe landing. The AP-3DTM solution, proposed by AGNC, comprises the following major innovations: a) use for the first time of the theory of stereo-perspective for 3D landing models generation in real-time from non-calibrated passive imaging; b) unique landing prediction multi-sensor architecture based on AGNC’s robust Kalman filter. The Phase I research will result in identification and feasibility verification of the AP-3DTM’s most challenging components. Phase II will culminate in a fully functional working prototype.