摘要
随着机器视觉的广泛应用,三维重建技术越来越受关注,而运动恢复结构法(Structure From Motion,SFM)凭借其设备成本低、方便携带、适用场景广泛等优点成为三维重建的研究热点。文章从相机成像基础理论出发,介绍了坐标系、相机成像几何关系,然后将SFM问题的转化为数学问题进行SFM算法原理介绍。在MATLAB软件平台实现了增量SFM算法的应用,并得到了相机位姿和三维点云,为开展SFM应用研究提供新思路。
With the wide application of machine vision, 3D reconstruction technology has attracted more and more attention.Structure From Motion(SFM) method has become a research hotspot of 3D reconstruction due to its advantages of low equipment cost,convenient portability and wide range of applicable scenes. Starting from the basic theory of camera imaging, this paper introduces the coordinate system and the geometric relationship of camera imaging, and then transforms the SFM problem into a mathematical problem to introduce the principle of SFM algorithm. The incremental SFM algorithm is applied in MATLAB software platform, and the camera pose and 3 D point cloud are obtained, which provides a new idea for SFM application research.
作者
李美燕
黄世玲
LI Meiyan;HUANG Shiling(Gaobo Software College,Nanning University,Nanning Guangxi 530200,China;College of Mechanical and ElectricalEngineering,Nanning University,Nanning Guangxi 530200,China)
出处
《信息与电脑》
2021年第18期30-33,共4页
Information & Computer
基金
南宁学院科研项目“基于改进SFM方法的航空摄影测量级城市三维重建应用研究”(项目编号:2019XJ23)
南宁学院教授培育项目“动力锂电池组管理系统技术及SOC估计策略的研究”(项目编号:2018JSGC17)。
