加权最小二乘优化倾斜摄影测量模型矢量化精度方法研究

Research on weighted least squares optimizing the accuracy of vectorization of oblique photogrammetry

无人机倾斜摄影测量技术以高精度、高效率、大范围的方式全面感知复杂场景,通过专业的数据采集设备和建模软件恢复三维场景,再进行量测。市场上的三维测图软件大部分是基于立面散点正交原理,不能有效地减弱人工立体测图产生的误差。本文提出一种加权最小二乘直线拟合优化倾斜摄影测量模型矢量化精度的方法,选取南昌市桑海管理委员会部分区域进行地形图数据采集和精度评定,试验表明,采用本文方法所建立的立体测图模型精度提高了2 cm。

UAV tilt photogrammetry technology comprehensively perceives complex scenes in a high-precision,high-efficiency,and large-scale manner,which recovers the three-dimensional scene through professional data acquisition equipment and modeling software,and then performs the measurement.Most of the three-dimensional mapping software on the market is based on orthogonal principle of facade scattered points,which cannot effectively reduce the errors caused by manual stereo mapping.This paper proposes a method of weighted least squares line to optimize the vectorization precision of the tilt photogrammetry model.Some areas of Nanchang Sanghai Management Committee of Nanchang city are selected for topographic map data collection and accuracy assessment in this paper.Experiments show that the accuracy of the stereo mapping model is improved by two centimeters through the method.