摘要
对无人机视觉导航的现有算法从视觉处理和状态估计与数据融合两大方面作了分类总结,重点讨论了无人机获取图像的特征提取、视觉传感器同其他传感器的数据融合、以及实时得到状态估计的各种方法,同时,也涉及了未知环境下地图的建立,最后指出了无人机视觉导航在实际应用中面临的主要问题。
In this paper, some existing algorithms concerning vision-based UAV navigation were classified and summarized from two aspects, vision processing and states estimation companied with data fusion. Feature extraction from pictures obtained by UAV, fusing camera with other sensors and many kinds of methods to obtain states estimation in real time were mainly discussed. Meanwhile, mapping the environment was also involved especially in unknown environment. At last, the main problems in practice associated with vision-based UAV navigation were present.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2010年第A01期62-65,共4页
Journal of System Simulation
基金
国家自然科学基金(60732064)
关键词
无人机
视觉导航
特征提取
数据融合
UAV
vision-based navigation
feature extraction
data fusion