摘要
提出了一种基于单目视觉/INS组合导航的无人机自主着陆末端导航信息提取的新方法;它只需要检测跑道上的两个特征点再结合?导系统的姿态角信息,就可以直接解算出飞机相对于机场的位置矢量;由于直线具有明显的可观性,所以对于特征点较难提取的可以通过对两条跑道边缘线和一条着陆阈值线进行检测,然后取阈值线与跑道边缘线的交点作为特征点;依据成像的几何原理,对算法进行了详细的推导,最后通过着陆实验表明该方法的提取与转换是切实可行的。
A new extraction method of vision--based/INS integrated navigation for UAV autonomous landing is proposed. It can calculate directly the position vector only by extracting two feature points on the runway and combining with the attitude information of INS. Due to lines are obviously visual, it can extracts the two edge lines and the threshold line on the runway if the feature points extracts difficultly, then put the nodal of the threshold line with the edge of the runway as feature points. Algorithm is given based on imaging geometry. It shows that the extraction and conversion are visible through the experience.
出处
《计算机测量与控制》
CSCD
北大核心
2009年第2期329-331,共3页
Computer Measurement &Control
基金
总装预研基金(9140A25040307HK0306)
关键词
计算机视觉
着陆导航
特征点
computer vision
landing navigation
feature point