摘要
文中针对倾斜摄影技术在港口三维重建中存在技术缺陷,提出一种无人机倾斜摄影与近地面摄影测量相结合的建模方法构建虚拟港口三维场景.对于多源点云数据,利用体素滤波下采样进行精简;基于采样一致性初始配准算法的点云粗配准和基于KD树改进Point-to-Plane ICP(Iterative Closest Point)算法的精配准实现密集点云数据融合;最后利用融合后的点云数据进行三维重建.结果表明:该方法较倾斜摄影建模在模型质量和精度上有大幅提升,模型误差符合I级产品标准,可用于构建虚拟港口三维场景.
Aiming at the technical defects of tilt photography technology in port 3D reconstruction,a modeling method combining UAV tilt photography with near-surface photogrammetry was proposed to construct a virtual port 3D scene.For multi-source point cloud data,firstly,voxel filtering was used for downsampling,and then dense point cloud data fusion was realized by coarse registration based on sampling consistency initial registration algorithm and fine registration based on KD-tree improved point-to-plane ICP(iterative closest point)algorithm.Finally,the fused point cloud data was used for 3D reconstruction.The research shows that this method has greatly improved the quality and accuracy of the model compared with oblique photography modeling,and the model error meets the I-class product standard,so it can be used to construct the virtual port three-dimensional scene.
作者
陈立家
王莹涛
王凯
吴小红
戚建祥
CHEN Lijia;WANG Yingtao;WANG Kai;WU Xiaohong;QI Jianxiang(School of Shipping,Wuhan University of Technology,Wuhan 430063,China;Hubei Key Laboratory of Inland Water Navigation Technology,Wuhan 430063,China;Zhejiang International Marine College,Zhoushan 316021,China)
出处
《武汉理工大学学报（交通科学与工程版）》
2024年第6期1237-1243,1250,共8页
Journal of Wuhan University of Technology（Transportation Science & Engineering）
基金
国家重点研发计划项目(2019YFB1600603)。
关键词
虚拟港口
点云
多源数据融合
三维重建
virtual port
point cloud
multi-source data fusion
3D reconstruction
