融合视惯RTK的非接触高精度点位测量方法  

Non-contact high-precision point positioning measurement method fusion with visual/IMU/RTK

作　　者：裴凌[1] 邓怡琳 李涛 刘国庆[1] 计洁 王超 PEI Ling;DENG Yilin;LI Tao;LIU Guoqing;JI Jie;WANG Chao(Shanghai Key Laboratory of Navigation and Location-based Services,Shanghai Jiao Tong University,Shanghai 200240,China;College of Information Engineering,Zhejiang University of Technology,Hangzhou 310023,China;Shanghai Huace Navigation Technology Ltd.,Shanghai 201706,China)

机构地区：[1]上海交通大学上海市北斗导航与位置服务重点实验室,上海200240 [2]浙江工业大学信息工程学院,杭州310023 [3]上海华测导航技术股份有限公司,上海201706

出　　处：《中国惯性技术学报》2024年第7期663-670,680,共9页Journal of Chinese Inertial Technology

基　　金：国家自然科学基金(62273229);上海市协同创新项目(HCXBCY-2023-020)。

摘　　要：为解决信号受遮挡或物理不可达区域的高精度点位测量问题,提出了一种融合视惯和实时差分定位(RTK)信息的非接触高精度点位测量方法。首先,利用由RTK测定的相机位置信息,在增量式重建中将相机位姿和特征点空间位置配准到全球地理坐标系下;其次,在全局优化中引入视觉重投影误差、RTK测定的相机位置、惯性测量单元测定的相机姿态作为约束,以提升特征点空间位置的测量精度;最后,将图像中的特征点坐标从像平面坐标系映射到全球地理坐标系下,实现对场景中特征点位坐标的非接触性测量。与OpenMVG的对比实验表明:所提方法在3m和5m距离处的非接触性点位测量精度分别提高了87.25%和75.88%,验证了其在非接触性点位测量问题上的高精度。To tackle high-precision point measurement challenges in signal-obstructed or inaccessible areas,a non-contact high-precision point positioning measurement method combining vision,inertial measurement unit(IMU)and real time kinematic(RTK)information is proposed.Firstly,utilizing the camera position information determined by RTK,the spatial positions of feature points and camera poses are registered to the geographical coordinate system during the incremental reconstruction process.Secondly,global optimization is introduced with constraints on visual reprojection error,RTK-determined camera positions,and IMU-determined camera attitudes to enhance the estimation accuracy of feature point spatial positions.Finally,the coordinates of the feature points in the image are mapped from the image plane coordinate system to the geographical coordinate system to achieve non-contact measurement of the spatial coordinates of the feature points in the scene.Comparative experiments with the point measurement method based on OpenMVG showed that the proposed method enhances point measurement accuracy by 87.25%at 3 m and 75.88%at 5 m,underscoring its superior accuracy in point position estimation.

关 键 词：点位测量 视觉/IMU/RTK 光束平差法 增量式运动恢复结构算法 

分 类 号：P23[天文地球—摄影测量与遥感]