中文摘要：

针对无人机（UAV）着陆时机身相对于着陆点的精确位姿参数的需求,设计了一套基于日盲区紫外成像的无人机自主着陆引导系统。利用日盲区紫外成像技术对典型特征的紫外信标点进行成像,基于摄像机透视投影变换以及直接线性变换求得位姿参数初值,并对初值进行非线性优化得到精确解,由此实现无人机自主着陆引导工作。利用日盲区紫外在近地面唯一性的特性,能消除自然背景光的影响,在低能见度条件下的优势尤其明显。经实验验证,系统距离定位相对精度能够达到0.5%,角度定位绝对精度优于1°,能够满足无人机着陆引导的精度要求。

英文摘要：

An unmanned aerial vehicle（UAV） landing guidance system based on solar-blind ultraviolet imaging is designed to meet the demand of precise pose parameters of fuselage to the landing sites. The system utilizes a solar- blind UV imaging system with dual channel and a UV beacon consisted of several sources, which is imaged by the uhraviolet-channel imaging system, and the spot position of UV beacon is extracted based on light intensity. Camera perspective projection model and direct linear transformation algorithm are adopted to solve the initial pose parameters. The accurate pose parameters of UAV to the UV beacon are obtained by the iterative optimization, which guides UV safely landing. Because of the uniqueness of solar-blind ultraviolet near the ground, the advantages of system in poor visibility conditions are particularly obvious. Experimental results show that the relative accuracy of range location can reach 0.5 % and the absolute accuracy of angle location is better than 1%. The proposed system can meet the requirements of UAV landing guidance.