Vision-based Quadrotor Micro-UAV Position and Yaw Estimation and Control

摘要

This paper describes a novel vision-based position and yaw estimation strategy for quadrotor micro-UAV control for target tracking based on identifying and tracking two points on a target using optic flows. It uses an optimization algorithm and Kalman filters for estimation of quadrotors position and yaw relative to the target. Further position, altitude and heading control of quadrotor are achieved based on vision sensing alone. In addition, experimental results are presented to validate performance of proposed approaches.