极坐标下的多源数据联合空三方法研究
A Polar Coordinate Based Combined Aerial Triangulation Method with Multi-source Data
尹峻 1王林 1李犁 2刘文轩 3李小凯3
作者信息
- 1. 合肥市测绘设计研究院有限公司,安徽 合肥 230061;合肥市地理信息与智慧城市工程技术研究中心,安徽 合肥 230061
- 2. 合肥巢湖空间测绘科技有限公司,安徽 合肥 230061
- 3. 武汉大势智慧科技有限公司,湖北 武汉 430223;自然资源部实景三维建设与城市精细化治理工程技术创新中心,湖北 武汉 430223
- 折叠
摘要
无人机倾斜摄影测量和三维激光扫描是构建实景三维中国的重要方法,高效且精细的城市级三维模型重建是新型基础测绘和智慧城市研究的热点问题.本文提出一种极坐标视觉框架下无人机倾斜影像、车载激光点云和机载激光点云的联合空三模型,以此为基础构建了一种多源数据支持的精细化城市级实景三维建模方法.实验结果表明,极坐标视觉下影像与点云联合平差模型可以有效提升空三处理精度,极大地改善了倾斜模型的建模质量,为多源数据融合的大规模城市级实景三维建设提供了一种新的实施路径.
Abstract
Unmanned aerial vehicle ( UAV) oblique photogrammetry and 3D laser scanning technology are fundamental methodologies for building realistic 3D China. The reconstruction of urban-level 3D models with high efficiency and accuracy is a focal point of the contempora-ry basic surveying and smart city research. In this paper,we propose a joint aerial triangulation model that integrates oblique imagery,vehicu-lar,and airborne laser point clouds in a polar coordinate visual framework,and then construct a refined real-world 3D modeling framework supported by multi-source data. Experimental results show that the combined aerial triangulation model of image and point clouds based on polar coordinate vision can effectively improve the accuracy of aerial triangulation process. It provides a new pathway to enhance the generation of realistic 3D city models with multi-source data fusion.
关键词
实景三维/联合空三/无人机倾斜影像/激光雷达点云Key words
3D real scene/combined aerial triangulation/UAV images/LiDAR point clouds引用本文复制引用
出版年
2024
NETL
NSTL
万方数据