Line Segment Matching Based on Geometric Constraints and Gradient Descriptor

doi:10.13474/j.cnki.11-2246.2018.0139

Abstract

Abstract:

Aiming at the reliability of line segment matching under the condition of non-constraint,this paper proposed a line matching algorithm based on geometric constraints and gradient descriptor.Based on the corresponding points and straight line extraction results,the algorithm firstly determined the candidate lines by corresponding triangulations and epipolar constraints.Secondly,it determined the overlapping of the candidate line which corresponding the reference line based on the epipolar of reference line endpoints.Thirdly,based on corresponding overlapping segments,it established straight line parallel support region for reference line and candidate line respectively,and then constructed gradient magnitude mean-standard deviation descriptor with Gaussian convolution.Finally,the corresponding line was determined by Euclidean distance in descriptors and nearest neighbor distance ratio.Experimental studies are carried out with different kinds of images,and the results show that the proposed method is reliable for line matching.

Key words: gradient magnitude descriptor, triangulation constraint, epipolar constraint, straight line parallel support region, nearest neighbor distance ratio

CLC Number:

P237

WANG Jingxue, ZHANG Xue. Line Segment Matching Based on Geometric Constraints and Gradient Descriptor[J]. 测绘通报, 2018, 0(5): 29-34.

References

[1] 王竞雪,宋伟东,韩丹,等.边缘视差连续性约束的航空影像特征线匹配算法[J].信号处理,2015,31(3):364-371.
[2] CHEN M,SHAO Z,LIU C,et al.Scale and Rotation Robust Line-based Matching for High Resolution Images[J].Optik-International for Light and Electron Optics,2013,22(124):5318-5322.
[3] HAN W,ZHENG J B.Fast and Accurate Stereo Matching Algorithm Based on Epipolar Line Restriction[J].Computer Engineering & Applications,2008,44(1):51-53.
[4] 朱红,宋伟东,杜北,等.重合度约束的近景影像直线匹配算法[J].信号处理,2015,31(8):912-917.
[5] 吴军,姚泽鑫,程门门.融合SIFT与SGM的倾斜航空影像密集匹配[J].遥感学报,2015,19(3):431-442.
[6] 朱庆,吴波,赵杰.基于自适应三角形约束的可靠影像匹配方法[J].计算机学报,2005,28(10):1734-1739.
[7] 张云生,朱庆,吴波,等.一种基于三角网约束的立体影像线特征多级匹配方法[J].武汉大学学报(信息科学版),2013,38(5):522-527.
[8] WO B,ZHANG Y S,ZHU Q.Integrated Point and Edge Matching on Poor Textural Images Constrained by Self-adaptive Triangulations[J].ISPRS Journal of Photogram-metry and Remote Sensing,2012,68(1):40-55.
[9] FAN B,WU F,HU Z.Line Matching Leveraged by Point Correspondences[C]//IEEE Conference on Computer Vision and Pattern Recognition.San Francisco:IEEE,2010:390-397.
[10] FAN B,WU F,HU Z.Robust Line Matching through Line Point Invariants[J].Pattern Recognition,2012,45(2):794-805.
[11] 陆丽娜.基于三焦点张量的多视图目标三维重建[D].南京:南京航空航天大学,2010:86-88.
[12] 娄安颖,宋伟东,刘薇.基于单应矩阵的直线匹配[J].遥感信息,2011,15(3):9-13.
[13] 易成涛,王孝通,徐晓刚.基于极线约束的角点匹配快速算法[J].系统仿真学报,2008,20(S1):371-374.
[14] 陈浩,高志强,密保秀,等.引入极线约束的SURF特征匹配算法[J].中国图象图形学报,2016,21(8):1048-1056.
[15] SCHMID C,ZISSERMAN A.The Geometry and Matching of Lines and Curves over Multiple View[J].International Journal of Computer Vision,2000,40(3):199-233.
[16] 梁艳,盛业华,张卡,等.利用局部仿射不变及核线约束的近景影像直线特征匹配[J].武汉大学学报(信息科学版),2014,39(2):229-233.
[17] LOWE D G.Distinctive Image Features from Scale-invariant Keypoints[J].International Journal of Computer Vision,2004,60(2):91-110.
[18] 缪君,储珺,张桂梅.一种仿射不变的直线描述子与直线匹配[J].电子学报,2015,43(12):2505-2512.
[19] 王志衡,吴福朝.均值-标准差描述子与直线匹配[J].模式识别与人工智能,2009,22(1):32-39.