作者: Wenguang Mao , Zaiwei Zhang , Lili Qiu , Jian He , Yuchen Cui
关键词: Control theory 、 Drone 、 Orientation (computer vision) 、 Computer science 、 Model predictive control 、 Global Positioning System 、 Simulation 、 Mobile phone 、 Metric (mathematics) 、 Real-time computing 、 Jerk
摘要: With the availability of inexpensive and powerful drones, it is possible to let drones automatically follow a user for video taping. This can not only reduce cost, but also support taping in situations where otherwise (e.g., during private moments or at inconvenient locations like indoor rock climbing). While there have been many follow-me on market outdoors, which rely GPS, enabling function more challenging due lack an effective approach track users environments. To this end, we develop holistic system that lets mobile phone carried by accurately drone's relative location control maintain specified distance orientation automatic We series techniques (i) using acoustic signals with sub-centimeter errors even under strong propeller noise from drone complicated multipath environments, (ii) solve practical challenges applying model predictive (MPC) framework drone. The latter consists developing measurement-based flight models, designing measurement provide feedback controller, predicting user's movement. implement our AR Drone 2.0 Samsung S7. extensive evaluation shows effectively following within 2-3 cm 1-3 degree errors, respectively. videos taped are smooth according jerk metric.