›› 2013, Vol. 42 ›› Issue (5): 691-698.

• 学术论文 • 上一篇    下一篇

多匹配基元集成的多视影像密集匹配方法

王竞雪1,朱庆2,王伟玺3   

  1. 1. 辽宁工程技术大学
    2. 西南交通大学
    3. 深圳市规划国土发展研究中心
  • 收稿日期:2012-09-13 修回日期:2013-12-04 出版日期:2013-10-20 发布日期:2014-01-23
  • 通讯作者: 王竞雪 E-mail:xiaoxue1861@163.com
  • 基金资助:
    国家自然科学基金面上项目;国家自然科学基金项目;武汉大学测绘遥感信息工程国家重点实验室开发基金;高等学校博士学科点专项科研基金资助课题

A Dense Matching Algorithm of Multi-View Image Based on the Integrated Multiple Matching Primitives

  • Received:2012-09-13 Revised:2013-12-04 Online:2013-10-20 Published:2014-01-23

摘要: 针对多视影像密集匹配中的遮挡问题,本文提出一种像方特征点和物方平面元集成的多视影像密集匹配方法。该方法利用规则格网划分的空间平面作为基础,对两种不同形式的匹配基元进行集成。首先通过多视影像上特征点投影范围和平面元位置相互制约,实现特征点和平面元的同时匹配,为后续匹配提供初始可靠的DSM;然后在此基础上,结合铅垂线轨迹法和基于高度遮挡检测方法对平面上规则分布的平面元进行匹配,加密初始匹配结果。最后通过对某地区四张UCX数字航空影像的匹配试验,验证了本文方法的有效性。

关键词: 多视影像匹配, 像方特征点, 物方平面元, 遮挡检测, 铅垂线轨迹法

Abstract: Focusing on the occlusion problem in multi-view image matching, this paper presents one new dense matching algorithm for multi-view image by integrating the feature points in image object and plane elements in the space object, which are based on the space plane divided in regular grids. Firstly, it realizes the simultaneously matching of feature points and plane elements by the constrains of the projection ranges of feature points in multi-view image and the positions of plane elements, and provides a reliable initial DSM for the next dense matching; Then it carries on the dense matching of regular distribution plane elements by combining the Vertical Line Locus method and the Height-based Occlusion Detection method, and densifies the initial matching results; Finally, the validity of the algorithm proposed in this paper is verified by the experiments using four UltraCamX (UCX) digital aerial images.

Key words: multi-view image matching, feature point in image object, plane element in space object, occlusion detection, Vertical Line Locus

中图分类号: