融合倾斜摄影和激光点云的城市实景三维重建方法  

3D reconstruction method of urban reality by integrating oblique photography and laser point cloud

作　　者：贾晓博 JIA Xiaobo(Anhui Surveying and Mapping Technology Training Center,Hefei,Anhui 230031,China)

机构地区：[1]安徽省测绘技术培训中心,安徽合肥230031

出　　处：《北京测绘》2025年第2期170-176,共7页Beijing Surveying and Mapping

摘　　要：为改善城市实景模型构建效果,为智慧城市建设提供数据支持,本研究提出基于倾斜摄影和激光点云融合的城市实景三维重建方法。通过将不同站点激光点云数据以及倾斜影像提取的点云数据转换到目标坐标系,完成多源点云的初步匹配,再利用引入旋转角约束以及动态迭代系数的改进迭代最近点算法对其作精确配准,从而完成多源点云融合。在Context Capture软件中,对融合后的点云数据进行处理,生成三维表面的三角网格表示。通过对三角网格的优化调整,得到三角网格模型,为模型添加纹理贴图,从而成功实现城市实景的三维重建。实验结果表明:重建后的城市实景三维模型在纹理细节上表现突出、形态逼真,与原始点云数据之间存在极小的点位误差,范围在0.005~0.017 m。To improve the construction effect of urban reality models and provide data support for smart city construction,a three-dimensional(3D)reconstruction method of urban reality based on oblique photography and laser point cloud fusion was proposed.After converting laser point cloud data from different stations and point cloud data extracted from oblique images to the target coordinate system,preliminary matching of multi-source point clouds was completed.Then,an improved iterative nearest point algorithm with rotation angle constraints and dynamic iteration coefficients was used for accurate registration,thus completing multi-source point cloud fusion.In the Context Capture software,the fused point cloud data was processed to generate a triangular mesh representation of the 3D surface.After optimizing and adjusting the triangular mesh,a triangular mesh model was obtained,and texture maps were added to the model,successfully achieving 3D reconstruction of urban reality.The experimental results show that the reconstructed 3D model of urban reality exhibits outstanding texture details and realistic shapes,with extremely small point position errors between the original point cloud data and the reconstructed model ranging from 0.005-0.017 m.

关 键 词：倾斜摄影 激光点云 三维重建 迭代最近点算法 三角网格模型 旋转角约束 

分 类 号：P235[天文地球—摄影测量与遥感]