摘要
针对传统空中三角(简称空三)测量方法处理倾斜航空影像的不足,提出了一种基于SFM算法的倾斜航空影像空三测量方法。首先利用垂直影像空三结果获取初始区域网模型,垂直影像空三解算可通过常规的空三算法或现有的软件获取。然后采用SFM算法,在初始区域网模型的基础上,递增式地恢复倾斜影像的外方位元素,最后将垂直影像和倾斜影像联合进行绝对定向、区域网光束法平差。选用玉树数据进行空三试验,结果表明:本文提出的SFM方法能够精确解算出倾斜航空影像的外方位元素,且空三结果优于现有倾斜影像空三软件Pix4Dmapper,可以作为一种倾斜航空影像定向方法。
This paper proposed on an aerial triangulation method for oblique aerial images based on SFM algorithm.The initial regional network model from the orientated vertical images can obtain by conventional aerial triangulation methods or by existing aerial triangulate software.Then,put the initial regional network model into SFM algorithm and gradually recover the elements of exterior orientation for oblique images.Experiment was conducted on Yushu oblique images.The experimental results show that SFM algorithm can precisely calculate the elements of exterior orientation for oblique images even better than the existing aerial triangulate software Pix4 Dmapper.
出处
《测绘科学》
CSCD
北大核心
2017年第11期97-101,共5页
Science of Surveying and Mapping
基金
国家科技支撑计划项目(2012BAJ23B03)
关键词
倾斜航空影像空三
运动估计结构
区域网模型
外方位元素
aerial triangulation of oblique aerial images
SFM（struct from motion）
regional networkmodel
elements of exterior orientation
