Due to the limitations and power requirements of modern computational hardware and sensors, the central goals of smallness and high-level autonomy (usually associated with higher algorithmic complexity and computational cost) are directly at odds with each other. Current small UAVs often shed weight through the use of limited sensors and rudimentary algorithms, while very high-level autonomy usually comes packaged in a UAV that is by no means small. Our proposed project is uniquely situated to tackle this highly autonomous small UAV challenge through the innovative combination of high-performance, compact hardware and efficient, advanced algorithms. The proposing team will combine a flight-proven, state-of-the-art miniature avionics system, flight-proven, real-time image processing techniques, and efficient, high-level autonomous guidance algorithms to develop a Small, Image-aided Navigation and Autonomous Path-planning System (SINAPS) for small UAVs. The Phase I result will be a complete flight-proven SINAPS with very high levels of autonomy contained in an extremely small, lightweight package.
Keywords: Uav, Image-Aided Navigation, Autonomy, Collision Avoidance, High Speed Maneuvering