基于视觉的飞机自主着陆导航  

Vision-based aircraft autonomous landing navigation

作　　者：赵昊昱[1] 李红[1] 彭嘉雄[2] 

机构地区：[1]华中科技大学数学系,湖北武汉430074 [2]华中科技大学图像识别与人工智能研究所,湖北武汉430074

出　　处：《系统工程与电子技术》2007年第7期1131-1133,共3页Systems Engineering and Electronics

基　　金：国家自然科学基金项目资助课题(10477007)

摘　　要：提出一种基于视觉的飞机自主着陆导航信息提取新方法。它克服了国外方法中利用点对应关系而难于提取特征点的问题,只需对两条跑道边线和一条着陆阈值线进行检测,这些线具有明显的可视性和直线性。由此可直接算出飞机的姿态向量,若跑道的宽度已知,还可以算出飞机的位置矢量。依据成像的几何原理,文中对算法进行了完整的推导,通过对飞机着陆过程中实际采集的跑道图像序列进行处理和实验,表明基于视觉的飞机自主着陆导航信息提取与转换是可行的。A new extraction method of vision-based navigation information for aircraft autonomous landing is propsed. It avoids the difficulty of feature points extraction in oversea method useing points correspondence relation. It extracts the two edge lines and the threshold line on the runway only, These lines are obviously visual and straight Then the attitude vector of aircraft is directly calculated. And the position vector of aircraft is computed too if the runway width is known. A complete drive of the algorithm is given based on imaging geometry. The processing and experience on a runway image sequence acquired in a real aircraft landing are performed. It shows that the extraction and conversion of vision based navigation information for aircraft autonomous landing are feasible.

关 键 词：机器视觉 着陆导航 射影几何 消失点 直线检测 

分 类 号：V249[航空宇航科学与技术—飞行器设计]