模拟地形辅助的无人机倾斜摄影测量图像匹配  

Image matching for UAV oblique photogrammetry assistant by simulated terrain

作　　者：于大鹏 梁玉斌 冯晨阳 邱炎 YU Dapeng;LIANG Yubin;FENG Chenyang;QIU Yan(Overseas Company,Shen Kan Engineering&Technology Corporation,Shenyang 110169,China;School of Geographic and Environ-mental Sciences,Tianjin Normal University,Tianjin 300387,China;Geological Environment Information Center,Tianjin Institute of Geological Survey,Tianjin 300191,China)

机构地区：[1]中冶沈勘工程技术有限公司海外分公司,沈阳110169 [2]天津师范大学地理与环境科学学院,天津300387 [3]天津市地质调查研究院地质环境信息中心,天津300191

出　　处：《天津师范大学学报（自然科学版）》2020年第4期64-70,共7页Journal of Tianjin Normal University：Natural Science Edition

基　　金：天津师范大学市级工程中心开放研究基金资助项目(YF11700103)。

摘　　要：针对传统无人机摄影测量中基于位置的图像匹配方法存在的误匹配和不充分匹配问题,提出一种模拟地形辅助的无人机倾斜摄影测量图像匹配方法.该方法通过程序模拟生成测区的地面采样点集,利用采样点集和无人机POS数据,基于严格的成像模型确定影像间准确的匹配关系,并根据准确的匹配关系实现图像匹配和POS辅助全自动空中三角测量.实验结果表明:本研究提出的模拟地形辅助图像匹配方法解决了基于位置的图像匹配方法引起的误匹配和不充分匹配问题,使倾斜影像定向的成功率提高至100%,将倾斜影像的POS辅助全自动空三定位精度提高了1.4~3.6倍.Aiming at the problems of mismatch and insufficient matching of the location-based image matching method in the traditional unmanned aerial vehicle(UAV)photogrammetry,a simulated terrain assisted image matching method for UAV oblique photogrammetry was proposed.This method generates the ground sampling point set in the measurement area by the simulation program,and the accurate matching relationship between images was determined based on the strict imaging model using sampling point set and the drone POS data.The image matching and POS-assisted automatic aerial triangulation were realized based on the accurate matching relationship.Experimental results show that the simulated terrain-assisted image matching method proposed in this study solves the problem of mismatch and insufficient matching caused by the location-based image matching method.It improves the success rate of oblique image orientation to 100%,and the positioning accuracy of the POS-assisted automatic aerial triangulation increases by 1.4 to 3.6 times.

关 键 词：倾斜摄影测量 无人机 地形模拟 图像匹配 空中三角测量 

分 类 号：P237[天文地球—摄影测量与遥感]