Autonomous MWIR Guidance

This Phase I feasibility effort will formulate a comprehensive algorithm incorporating (a) multispectral/stereo sensor fusion, and (b) sensor/terrain database correlation, to generate a three-dimensional range estimate to relevant terrain and man-made objects of interests, to enhance situational awareness during approach, landing, and taxiing at a representative airport. It is also desirable that this navigational range estimate be suitable for usage in an automatic or manual vehicles guidance or control system. Consequently, the candidate algorithm must also be amenable to implementation and evaluation on a real-time information processing system during the subsequent Phase II effort. The proposed visible/IR sensor fusion algorithm will be based on selecting and correlating features that maximize luminance contrast of 11 x 11 (say) image pixel tiles. The proposed sensor/terrain (airport) database range correlation algorithm will be based on representing the terrain as a compact wireframe model and correlating the multispectral stereo sensor based imagery with the corresponding wireframe model based synthetic imagery extracted from the prestored airport database. The need for multispectral sensor fusion is driven by the wide dynamic range of the ambient illumination typical of outdoor imagery. The requirement for stereo sensors and sensor/database correlation is driven by the need to generate consistent and sufficiently precise range estimates, suitable for real time vehicle guidance and control applications. It is anticipated that this technology will be required by the planned national airspace plane, due to crew station visibility restrictions during approach and landing operations.