当前位置>主页 > 期刊在线 > 信息技术 >

信息技术22年20期

基于局部特征的自然场景配准算法比较分析
赵思雨,刘岩松,柳倩
(沈阳航空航天大学,辽宁 沈阳 110136)

摘  要:为研究 SIFT、SURF 和 ORB 三种局部特征配准算法在自然场景中的实时性和鲁棒性,对三种算法进行特征提取与匹配实验。通过比较同等配准条件下算法耗时分析各算法的实时性。通过对比图像在光照变换、几何变换、模糊变换和噪声变换下特征匹配数目分析各算法的鲁棒性。实验结果表明,ORB 算法具良好的实时性,SIFT 算法鲁棒性较强,SURF 算法的性能介于两者之间。


关键词:特征配准算法;特征提取;特征匹配;实时性;鲁棒性



DOI:10.19850/j.cnki.2096-4706.2022.20.001


基金项目:辽宁省教育厅青年科技人才“育苗”项目 (JYT2020130);痕迹检验鉴定技术公安部重点实验室开放课题(HJKF201907);公安部文件检验重点实验室开放课题 (FTKF202102)


中图分类号:TP317.4                                       文献标识码:A                                  文章编号:2096-4706(2022)20-0001-06


Comparison and Analysis of Natural Scene Registration Algorithms Based on Local Features

ZHAO Siyu, LIU Yansong, LIU Qian

(Shenyang Aerospace University, Shenyang 110136, China)

Abstract: In order to study the real-time and robustness of SIFT, SURF and ORB local feature registration algorithms in natural scenes, feature extraction and matching experiments are carried out on the three algorithms. The real-time performance of each algorithm is analyzed by comparing the time consuming of each algorithm under the same registration conditions. The robustness of each algorithm is analyzed by comparing the number of feature matching of images under illumination transformation, geometric transformation, fuzzy transformation and noise transformation. Experimental results show that ORB algorithm has good real-time performance, SIFT algorithm has strong robustness, and SURF algorithm has a performance between the two.

Keywords: feature registration algorithm; feature extraction; feature matching; real-time performance; robustness


参考文献:

[1] 黄敏 . 基于特征点的图像配准技术研究 [D]. 南昌:南昌大学,2021.

[2] LOWE D G. Distinctive Image Features from Scale-Invariant Keypoints [J].International Journal of Computer Vision,2004,60(2):91-110.

[3] BAY H,TUYTELAARS T,Gool L V. SURF:speeded up robust features [C]//ECCV’06:Proceedings of the 9th European conference on Computer Vision.Graz:Springer-Verlag,2006:404-417.[4] 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,2012:1423-1435.

[5] ROSTEN E,PORTER R,DRUMMOND T. Faster and Better:A Machine Learning Approach to Corner Detection [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2010,32(1):105-119.

[6] CANLONDER M,LEPETIT V,STRECHA C,et al. BRIEF:Binary Robust Independent Elementary Features [C]//Computer Vision - ECCV 2010.Heraklion:Spring,2010:778-792.

[7] 刘伟,钱莉 . 基于 OpenCV 环境的 SIFT、SURF、ORB 算法比较分析 [J]. 化工自动化及仪表,2018,45(9):714-716+721.

[8] 索春宝,杨东清,刘云鹏 . 多种角度比较 SIFT、SURF、BRISK、ORB、FREAK 算 法 [J]. 北京测绘,2014(4):23-26+22.

[9] 徐妍,王鹏辉,刘煜,等 . 基于 Matlab 的 SIFT 和 SURF算法在无人机影像配准中的对比研究 [J]. 矿山测量,2017,45(6):36-39.

[10] 林志东,杜滢钊 . 特定条件下 SIFT 与 SURF 图像匹配算法的性能对比 [J]. 厦门理工学院学报,2019,27(3):30-36.

[11] 陈敏,汤晓安 .SIFT 与 SURF 特征提取算法在图像匹配中的应用对比研究 [J]. 现代电子技术,2018,41(7):41-44.

[12] 查冰,张力,艾海滨 . 不同特征提取算法对相机运动估计的适用性研究 [J]. 测绘科学,2018,43(3):92-98.

[13] 倪嘉联,郭昱含,王强,等 . 不同场景下影像特征提取算法的适用性研究 [J]. 测绘与空间地理信息,2021,44(8):39-43.

[14] BEIS J S,LOWE D G. Shape indexing using approximate nearest-neighbor search in high-dimensional spaces [C]//Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern 

Recognition.San Juan:IEEE,1997:1000-1006.[15] FISCHLER M A,BOLLES R C. Random sample consensus:a paradigm for model fitting with applications to image analysis and automated cartography [J].Readings in Computer Vision, 1987:726-740.


作者简介:赵思雨(1998—),女,满族,辽宁沈阳人,硕士研究生在读,主要研究方向:交通信息工程及控制。刘岩松 (1963-),男,汉族,辽宁沈阳人,教授,博士,主要研究方向:交通运输工程 ;柳倩 (1984—),女,汉族,山西忻州人,博士,讲师。主要研究方向:交通运输规划与管理。