作者: Shaowu Yang , Sebastian A. Scherer , Andreas Zell
DOI: 10.1007/S10846-012-9749-7
关键词:
摘要: In this paper, we present an onboard monocular vision system for autonomous takeoff, hovering and landing of a Micro Aerial Vehicle (MAV). Since pose information with metric scale is critical flight MAV, novel solution to six degrees freedom (DOF) estimation. It based on single image typical pad which consists the letter "H" surrounded by circle. A algorithm robust real-time recognition implemented. Then 5 DOF estimated from elliptic projection circle using projective geometry. The remaining geometric ambiguity resolved incorporating gravity vector inertial measurement unit (IMU). last degree pose, yaw angle ellipse fitted "H". efficiency presented demonstrated comprehensively comparing it ground truth data provided tracking its estimates as control inputs flights quadrotor.