期刊文献+

基于相关点与几何约束的直线匹配方法 被引量:1

Related Points and Geometric Constraints Based Straight Line Matching Method
下载PDF
导出
摘要 由于直线本身缺乏显著特征,使用点与直线相关联的方法可提高直线特征的描述能力。针对特征点间及直线间均存在误匹配的问题,提出一种基于相关点与几何约束的直线匹配方法。首先,将直线邻域内满足条件的特征点定义为直线相关点,利用相关点与直线构造仿射不变量来描述直线,通过度量仿射不变量的相似性进行直线初匹配;然后计算仿射变换模型,通过直线角度和几何距离位置约束进行精匹配,去除误匹配直线,得到最终匹配结果。实验结果分析表明:本研究算法不仅适用于发生仿射变换的各种图像,得到更多的匹配直线对,匹配精度高,对发生仿射变换的图像、纹理复杂和纹理简单的图像,均能取得较好的直线匹配效果。 To combine related points and line can increase the line's features description ability. However,straight lines exist error matching situations. In order to improve line matching accuracy,a straight line matching method based on related points and geometric constraints is proposed. Firstly,related points are chosen and used to construct affine invariant to describe line feature. And lines are matched coarsely based on the similarity of affine invariant. Then,affine transformation model is calculated and straight lines are matched finely by the constraints of line angle and distance. The experimental results show that the method is suitable for image with affine transformation images,can obtain more matched lines pairs with high accuracy. It can obtain good line matching results for images with affine transformation and images with complex or simple texture.
作者 符祥 尹红 黄勇 曾接贤 FU Xiang;YIN Hong;HUANG Yong;ZENG Jie-xian(School of Software, Nanchang Hangkong University, Nanchang 330063, Chin)
出处 《南昌航空大学学报(自然科学版)》 CAS 2018年第1期41-47,共7页 Journal of Nanchang Hangkong University(Natural Sciences)
基金 国家自然科学基金(61662049 61763033 61741312 61762067) 国家留学基金委资助项目(201608360163)
关键词 SIFT特征点 仿射不变性 几何约束 直线匹配 straight line matching feature points affine invariants geometric constraints
  • 相关文献

参考文献5

二级参考文献53

  • 1周焰,李德仁,徐长勇.尺度变化对形状的比例不变特征的影响[J].华中科技大学学报(自然科学版),2003,31(3):11-13. 被引量:3
  • 2高广珠,李忠武,余理富,何智勇.归一化互相关系数在图像序列目标检测中的应用[J].计算机工程与科学,2005,27(3):38-40. 被引量:17
  • 3杨化超,邓喀中,张书毕.基于Hough变换的航空影像建筑物半自动提取[J].测绘科学,2006,31(6):93-94. 被引量:19
  • 4文贡坚.一种基于特征编组的直线立体匹配全局算法[J].软件学报,2006,17(12):2471-2484. 被引量:21
  • 5Tang A W K, Ng T P, Hung Y S, et al. Projective Reconstruction from Line-Correspondences in Multiple Uncalibrated Images. Pattern Recognition, 2006, 39 (5) : 889 - 896
  • 6Aider O A, Hoppenot P, Colle E. A Model-Based Method for Indoor Mobile Robot Localization Using Monocular Vision and Straight-Line Correspondences. Robotics and Autonomous Systems, 2005, 52(2/3 ) : 229 - 246
  • 7Shi Fanhuai, Wang Jianhua, Zhang Jing, et al. Motion Segmentation of Multiple Translation Objects from Line Correspondences. Pat- tern Recognition, 2005, 38(10) : 1775 -1778
  • 8Bartoli A, Sturm P. Multiple-View Structure and Motion from Line Correspondences // Proc of the IEEE International Conference on Computer Vision. Madison, USA, 2003 : 207 -212
  • 9Belongie S, Malik J, Puzicha J. Shape Matching and Object Recognition Using Shape Contexts. IEEE Trans on Pattern Analysis and Machine Intelligence, 2002, 24(4) : 509 -522
  • 10Lowe D G. Distinctive Image Features from Scale-Invariant Keypoints. International Journal of Computer Vision, 2004, 60 ( 2 ) : 91 -110

共引文献62

同被引文献7

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部