为了提高无人机着陆的自主性和安全性,基于视觉信息提出了一种自主着陆过程中无人机姿态和位置估计新方法。该方法将无人机的滚转、偏航和俯仰运动视为跑道做反方向运动,无人机视为静止,利用Hough变换检测跑道边界线并提取图像直线特征,同时融合无人机的平移运动和旋转旋动,利用地面直线和图像直线间对应关系,根据跑道边界线在图像中的直线特征从而估计出无人机姿态和位置。最后采用仿真实例来验证所研究的无人机姿态和位置估计算法的有效性。

In order to improve the autonomy and security of an Unmanned Aerial Vehicle (UAV),a vision-based method was proposed for attitude and position estimation of it during its autonomous landing. In this paper,the UAV’s roll,yaw and pitch motion are assumed as the runway’s retrograde motion while the UAV as static. The runway boundary is detected and straight line features are extracted by use of Hough transformation. Meanwhile,the rotary motion and translational motion are compounded. According to the relationship between straight line on the ground and in images,and straight line features of runway boundary in images,the estimation algorithm for UAV’s attitude and position is deduced. The simulation results of the example show the effective performance of the studied algorithm.