摘要
针对无人机在飞控信道被阻塞和失去卫星定位信号的情况,研究了依靠自主视觉导航来提高生存能力的可能。图像特征信息的获取和匹配是建立图像间的对应关系、实现视觉导航的关键,基于SIFT算法对无人机俯拍图像进行了局部特征的提取、描述和匹配研究,实验结果表明了SIFT算法在无人机俯拍图像的景象匹配中的不变性和准确性,证明了该算法在无人机视觉导航中的应用潜力。
In view of the case that the UAV （Unmanned Aerial Vehicles） flight control channel is blocked and satellite positioning signal is lost,the research is carried out to explore the possibility of enhancing survivability by autonomous visual navigation. The acquisition and matching of image characteristic information is essential to visual navigation and establishment of the corresponding relation between images.The research on feature extraction,description and matching of UAV top views is carried out based on the SIFT （ Seale-Invariant Feature Transform） algorithm.Test results indicate the invariance and accuracy of the SIFT algorithm used in the scene matching of UAV top views,which reveal the great application potential in UAV visual navigation.
出处
《无线电工程》
2017年第5期19-22,共4页
Radio Engineering
基金
中国博士后科学基金资助项目(2015M580217)
关键词
视觉导航
无人机
SIFT算法
特征提取
特征描述
特征匹配
visual navigation
UAV
SIFT algorithm
feature extraction
feature description
feature matching
