基于多重约束条件的线特征多级匹配方法

计算机科学 ›› 2014, Vol. 41 ›› Issue (Z11): 83-87.

基于多重约束条件的线特征多级匹配方法

李俊瑶,顾宏斌,孙瑾,王德志

南京航空航天大学民航学院南京210016;南京航空航天大学民航学院南京210016;南京航空航天大学飞行模拟与先进培训工程技术研究中心南京210016;南京航空航天大学民航学院南京210016;南京航空航天大学飞行模拟与先进培训工程技术研究中心南京210016;南京航空航天大学民航学院南京210016

出版日期:2018-11-14 发布日期:2018-11-14
基金资助:
本文受中央高校基本科研业务费专项资金(NS2013066)资助

Multi-lever Line Matching Method Based on Multiple Constraints

LI Jun-yao,GU Hong-bin,SUN Jin and WANG De-zhi

Online:2018-11-14 Published:2018-11-14

摘要/Abstract

摘要： 为解决直线断裂、遮挡以及图像对视差较大等造成的直线匹配困难,提出了一种基于多重约束条件的线特征多级匹配方法。该方法在已匹配的可靠种子点完成点-线匹配的基础上,基于单应性矩阵利用几何特征信息约束和核线约束完成同名直线搜索实现线-线匹配,最后提出一种自适应的直线相似性约束实现线-面匹配获得正确的同名直线。通过与现有直线匹配算法的对比实验表明,该方法提高了匹配精度,可以实现含有遮挡、直线断裂和视差较大的图像中的直线匹配。

关键词: 直线匹配,单应性矩阵,核线,多级匹配,自适应

Abstract: To match the images of larger parallax change,blocked or linear fracture,this paper proposed a multi-level line matching method based on and multiple constraints.Firstly,constraint of the linesegment in the neighborhood with matched reliable seed points to accomplish point-line matching is based on seed points;Next,perform of the line-line matching within geometric feature constraints is based on homography matrix and epipolar line constraint；Finally,we completed line matching with line-surface matching using a adaptive linear similarity constraint method.Experiments show that this method can accurately match images of larger issues,blocked or linear fracture and solve the problems that existing line matching algorithm has lower accuracy and unable to fulfil large parallax changing.

Key words: Line matching,Homography matrix,Epipolar line,Multi-level matching,Adaptation

李俊瑶,顾宏斌,孙瑾,王德志. 基于多重约束条件的线特征多级匹配方法[J]. 计算机科学, 2014, 41(Z11): 83-87. https://doi.org/

LI Jun-yao,GU Hong-bin,SUN Jin and WANG De-zhi. Multi-lever Line Matching Method Based on Multiple Constraints[J]. Computer Science, 2014, 41(Z11): 83-87. https://doi.org/

参考文献

[1] Tian Y,Gerke M,Vosselman G,et al.Automatic Edge Matching Across an Image Sequence Based on Reliable Points[C]∥International Archives of the Photogrammetry,Remote Sensing and Spatial Information Sciences.2008
[2] 寇媛,徐景中.基于结构特征的遥感影像匹配[J].中国图象图形学报,2013,18(5):565-573
[3] 郁理,郭立,袁红星.基于分级置信度传播的立体匹配新方法[J].中国图象图形学报,2011,16(1):103-109
[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,2006,7(12):2471-2484
[5] 黄良明,彭立中,等.利用几何代数进行线段模型匹配和运动估计[J].中国图象图形学报,2001,6(3):270-274
[6] Hartley R.A Linear Method for Reconstruction from Lines and Points[C]∥Proceddings of the 5th IEEE International Conference on Computer Vision.Cambridge,USA,1995
[7] Wilson R C,Hancock E R.Graph Matching with HierarchicalDiscrete Relaxation[J].Pattern Recognition Letters,1999,0(10):1041-1052
[8] Zhang M,Hou Y,Hu Z.Accurate Object Tracking Based onHomography Matrix[C]∥Computer Science & Service System (CSSS).2012 IEEE International Conference,2012:2310-2312
[9] Mcintosh J H,Mutch K M.Matching straight lines[J].CVGIP,1988(43):386-408
[10] Schmid C,Zisserman A.The geometry and matching of lines and curves over multiple views[J].International Journal of Compu-ter Vision,2000,40(3):199-233
[11] Yang L,Wang J,Chao S,et al.A study on the correlation coefficient matching algorithm of gene-ral digital stereo under the approximate epipolar line constraints[C]∥Multimedia Technology (ICMT).2011 IEEE International Conference,2011:954-957
[12] Liu Z,An J,Jing Y.A simple and robust feature point matching algorithm based on restricted spatial order constraints for aerial image registration[J].Geoscience and Remote Sensing,IEEE Transactions,2012,50(2):514-527
[13] Yang Y,Wang H,Liu B.A new stereo matching algorithmbased on adaptive window[C]∥Systems and Informatics (ICSAI).2012 International IEEE Conference,2012:1815-1819
[14] Smith S,Brady M.A new approach to low level image proces-sing[J].International Journal of Computer Vision,1997,23(1):45-78
[15] Zhang L.Automatic Digital Surface Model (DSM) Generation from Linear Array Images [D].ETH Zurich,Switzerland:Institute of Geodesy and Photogrammetry,2005
[16] Bay H,Ess A,Tuytelaars T,et al.Speeded-Up Robust Features(SURF)[J].Computer Vision and Image Understanding,2008,0(3):346-359

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed