无人驾驶飞机精确定位方法研究  

Research on the Locating of Remotely Piloted Vehicles

作　　者：邵永社[1,2] 陈鹰[1] 祝小平[3] 

机构地区：[1]同济大学测量与国土信息工程系遥感技术应用研究中心 [2]西北工业大学365研究所,西安710072 [3]西北工业大学365研究所

出　　处：《计算机工程与应用》2006年第23期188-191,共4页Computer Engineering and Applications

基　　金：国家863高技术研究发展计划资助项目(编号:2003AA133040)

摘　　要：GPS技术已广泛应用于无人驾驶飞机的空间定位和导航,但在战时、困难地区的应用会受到限制。文章提出了一种利用多点约束的影像匹配技术实现无人机精确定位的方法,该方法基于多点约束的思想提高了影像匹配的可靠性,同时利用匹配获取的多个同名点解决了由于无人机空中姿态变化引起的单点无法精确定位问题。该方法首先对无人机获取的实时影像和基准影像进行小波变换处理,以便从实时影像上获得足够的明显特征点,然后利用影像匹配和匹配点之间的位置约束关系获取基准影像上的同名点位,使用这些同名点,可以得到精确的无人机空间位置。GPS technique has been applied to locate and navigate the Remotely Piloted Vehicles（RPV） widely,but it is restricted in wartime and the difficult area.The paper presents a kind of method locating accurately the RPV using the image matching technique based on the restriction each other for the matching-points,which may improve the reliability and solve the problem that RPV can＇t been located accurately by the single matching-point for the RPV alterable attitude.The method firstly processes the image obtained by RPV and the reference image using wavelet transform so as to obtain the obvious feature points enough.Then the accurate coordinates of the matching points are calculated using the lease image matching and the restriction among the matching-points,which may calculate the accurate space location of the RPV.

关 键 词：无人机导航 空间定位 影像匹配 小波变换 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]