基于视觉导航的无人机自主精准降落系统

Autonomous and Precise Landing System of UAV Based on Visual Navigation

针对传统导航方式精度低、误差大等问题,结合视觉导航,设计了一种嵌套的着陆标识及识别的新算法。利用改进的ArUco检测算法快速识别着陆标志,通过融合四边形逼近算法和特征点信息得到无人机与降落标志的位姿,机载相机将其转换为导航信息传输给飞控板,实现无人机的自主精准降落。通过无人机室外实际应用实验,设计的降落标志及新型识别算法大大提高了无人机自主降落的精准性。

Aiming at the problems of low precision and large error of traditional navigation methods,combined with visual navigation,this paper designs a new algorithm for nested landing signs and identification.The improved ArUco detection algorithm is used to quickly identify the landing sign,and the pose of the UAV and the landing sign is obtained by fusing the quadrilateral approximation algorithm and feature point information and autonomous precision landing.Through the outdoor practical application experiment of UAV,the designed landing sign and new identification algorithm greatly improve the accuracy of UAV autonomous landing.