广东工业大学学报 ›› 2018, Vol. 35 ›› Issue (04): 56-60.doi: 10.12052/gdutxb.170166

• 综合研究 • 上一篇    下一篇

近景影像特征点匹配方法比较研究

余俊鹏, 林洁鸿, 詹松辉, 姚乃文   

  1. 广东工业大学 土木与交通工程学院, 广东 广州 510006
  • 收稿日期:2017-12-05 出版日期:2018-07-09 发布日期:2018-05-24
  • 作者简介:余俊鹏(1982-),男,高级工程师,博士,主要研究方向为摄影测量与遥感.
  • 基金资助:
    国家自然科学基金青年基金资助项目（41704019）；2017年大学生创新创业训练项目（201711845126）

A Comparative Study of Close-Range Image Feature Points Matching Methods

Yu Jun-peng, Lin Jie-hong, Zhan Song-hui, Yao Nai-wen   

  1. School of Transportation and Traffic Engineering, Guangdong University of Technology, Guangzhou 510006, China
  • Received:2017-12-05 Online:2018-07-09 Published:2018-05-24

HTML

PDF (PC)

2913

摘要： 近景摄影测量因其拍摄方式灵活，影像之间的相对几何变形大，常导致同名点匹配失败.本文采用SIFT、SURF、FAST+BRIEF和ORB 4种计算机视觉算法，对不同场景和摄影条件下的近景像对进行特征点检测与描述，结合BFMatch和FlannMatch两种方法对特征点实施匹配.实验表明，所用算法的计算耗时越长，匹配结果越好.SIFT、SURF适合于高精度连接点的自动生成，而FAST+BRIEF和ORB可用于相对几何变形小的立体影像密集点匹配.

关键词: 影像匹配, SIFT, SURF, 摄影测量

Abstract: Because of flexible shooting mode, the relative geometric deformation between images, one of the main problems of close-range photogrammetry is same name point matching. The SIFT, SURF, FAST + BRIEF, ORB were used to detect and describe the feature points of close-range images under different scenes and photography conditions. The BFMatch and FlannMatch methods were used to match the feature points. Experiments show that the longer the algorithm is, the better the matching result is. SIFT and SURF are both suitable for automatic generation of high-precision connection points. FAST + BRIEF and ORB can be used for stereo image matching with small relative geometric deformation.

Key words: image matching, SIFT, SURF, photogrammetry

中图分类号: 

  • P234.1
[1] 张岩,李建增,李德良,等. 基于特征的遥感图像匹配技术研究[J]. 无线电工程, 2016, 46(2):61-64.ZHANG Y, LI J Z, LI D L, et al. Research of remote sensing image matching based on key points[J]. Radio Engineering, 2016, 46(2):61-64.
[2] 王中杰. 基于图像特征提取和特征点描述的匹配方法的研究[J]. 电子设计工程, 2015, 23(24):173-174.WANG Z J. The research of matching method based on image for retraction and description of feature points[J]. Electronic Design Engineering, 2015, 23(24):173-174.
[3] 周莉莉,姜枫. 基于FAST和BRIEF的图像匹配算法[J]. 计算机工程与设计, 2015, 36(5):1269-1273.ZHOU L L, JIANG F. Image matching algorithm based on Fast and BRIEF[J]. Engineering and Design, 2015, 36(5):1269-1273.
[4] 冯莞舒. 基于特征点的影像匹配[J]. 电子测试, 2014, 303(17):17-19.FENG G S. Feature based image matching[J]. Electronic Test, 2014, 303(17):17-19.
[5] LOWE D G. Object recognition from local scale-invariant features[C]//IEEE International Conference on Computer Vision. Corfu:[s.n.], 1999:1150-1157.
[6] 毛星云,冷雪飞, 王碧辉,等. OpenCV3编程入门[M]. 北京:电子工业出版社,2016.
[7] BAY H, TINNE T, GOOL L V. Surf:speeded up robust features[J]. Proceedings of the European Conference on Computer Vision, 2006, 3951:404-417.
[8] ROSTEN E, DRUMMOND T. Machine learning for high-speed corner detection[J]. Proceedings of the European Conference on Computer Vision, 2006, 1:430-443.
[9] 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.KOU Y, XU J Z. Algorithm for remote sensing matching based on the structure characteristics[J]. Journal of Image and Graphics, 2013, 18(5):565-573.
[10] 王培珍,陈平,周芳,等. 复杂图像特征点提取与匹配方法[J]. 安徽工业大学学报(自然科学版), 2012, 29(1):73-77.WANG P Z, CHEN P, ZHOU F, et al. Method of feature extraction and matching for complex image[J]. Journal of Anhui University of Technology (Natural Science), 2012, 29(1):73-77.
[11] RUBLEE E, RABAUD V, KONOLIGE K, et al. ORB:an efficient alternative to SIFT or SURF[C]//2011 International Conference on Computer Vision. Barcelona:IEEE, 2011:2564-2571.
[12] HUANG W, WU L D, SONG H C. RBRIEF:a robust descriptor based on random binary comparisons[J]. Computer Vision Iet, 2013, 7(1):29-35.
[13] 彭丹妹,刘荣,刘洋. 基于OpenCV的图像特征点匹配方法的研究及实现[J]. 测绘与空间地理信息, 2016, 39(9):61-63.PENG D M, LIU R, LIU Y. Research and implementation of image feature point matching method based on OpenCV[J]. Geomatics & Spatial Information Technology, 2016, 39(9):61-63.
[14] HUO J,YANG N, CAO M Y, et al. A reliable algorithm for image matching based on SIFT[J]. Journal of Harbin Institute of Technology, 2012, 19(4):90-95.
[15] 余淮,杨文. 一种无人机航拍影像快速特征提取与匹配算法[J]. 电子与信息学报, 2016, 38(3):509-516.YU H, YANG W. A fast feature extraction and matching algorithm for unmanned aerial vehicle image[J]. Journal of Electronics & Information Technology, 2016, 38(3):509-516.
[16] 余俊鹏, 何炽荣. GPS-RTK辅助近景摄影测量光束法平差及精度分析[J]. 广东工业大学学报, 2017, 34(6):73-77.YU J P, HE C R. An accuracy analysis of GPS-RTK assisted close-range photogrammetric bundle adjustment[J]. Journal of Guangdong University of Technology, 2017, 34(6):73-77.
[1] 孟瑞, 廖开怀, 吴希文, 徐永明. 无人机测图的像控点布设优化及精度提升措施[J]. 广东工业大学学报, 2021, 38(02): 48-52,82.
[2] 叶志坚, 王福龙. 一种改进的FREAK算法的图像特征点匹配[J]. 广东工业大学学报, 2017, 34(06): 37-42.
[3] 余俊鹏, 何炽荣. GPS-RTK辅助近景摄影测量光束法平差及精度分析[J]. 广东工业大学学报, 2017, 34(06): 73-77.
[4] 李德隆, 刘伟. 基于改进的SIFT特征点的双目定位[J]. 广东工业大学学报, 2017, 34(01): 90-94.
[5] 雷禹, 何家峰. 基于改进SURF算法的SAR图像目标匹配[J]. 广东工业大学学报, 2014, 31(1): 65-69.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发