为了解决遥感图像配准的可靠性和实时性问题,提出了一种基于SURF算子的图像自动配准方法。SURF算法提取图像中适应旋转尺度变化的局部不变特征角点时使用积分图像的计算,可有效提高检测速度,用于对实时性要求高的场合,选择基于欧氏距离的最近邻与次近邻的距离之比作为相似度测量方法。多种不同场景图像对比实验结果表明,该方法可实现多种复杂图像的精确配准,自动化配准程度高。
Abstract
In order to solve the reliability and real-time problems of remote sensing image registration,an automatic image registration method based on SURF operator is proposed.SURF algorithm can effectively improve the detection speed by using integral image calculation when extracting locally invariant feature corners adapted to rotation scale changes.For situations with high real-time requirements,Euclian-based distance ratio of nearest neighbor to second-nearest neighbor is selected as the similarity measurement method.The experiment results show that this method can achieve accurate registration of complex images with high degree of automatic registration.
关键词
多传感器 /
遥感图像 /
SURF算子 /
图像配准 /
积分图像 /
RANSAC
Key words
multisensor /
remote sensing image /
SURF operator /
image registration /
integral image /
RANSAC
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 何文峰,查红彬.基于平面特征的深度图像配准[D].北京:北京大学,2006.
[2] L G Brown.A survey of image registration[J].ACM Computing Surveys,1992(24):326-376.
[3] B Zitová,J Flusser.Image registration methods:a survey[J].Image and Vision Computing,2003(21):977-1000.
[4] Tuytelaars T,Mikolajczyk K.Invariant feature detectots:a survey[J].Foundations and Trends in Computer Graphics and Vision,2019,3(3):177-280.
[5] 刘军,杨晶东.一种基于图像插值的SIFT算法研究[J].信息技术,2014,34(7):65-68.
[6] Yves Dufournaud.Matching Images with Different Resolutions.Theory in Computer Vision[J].Kluwer Academic Publishers,2004.
[7] 郑明玲,刘衡竹.基于特征提取和机器学习的医学图像分析[J].计算机学报,2011,27(9):1284-1289.
[8] Harris C,Stephens M.A combined comer and edge detector[C]//Proceedings of the 4 Alvey Vision Conference,1988.
[9] 邓传斌,郭雷,李维.基于SIFT的遥感图像配准方法[J].传感技术学报,2009,22(12):1742-1747.
[10] Bay H,Tuytelaars T,Cool L.SURF:Speeded Up Robust Features[C]//Proceedings of the 9 European Conference Computer Vision,2016.
[11] H Bay,T Tuytelaars,L Van Gool.Surf:Speeded up robust features[C]//European Conference on Computer Vision,2006.
[12] Paul Viola,Michael Jones.Rapid object detection using a boosted cascade of simple features[J].Computer Vision and Pattern Recognition,2018(I):511-518.
[13] Koenderink J.The structure of images[J].Biological Cybernetics 50,1984:363-370.
[14] Fischler M A,Bolles R C.Random sample conse- nsus:a paradigm for model fitting with applications to image automated cartography[J].Communications of the ACM,1981,24(6):381-395.
基金
*国家自然科学基金(52202491);西安交通工程学院中青年基金项目(2022KY-38)。
PDF(2274 KB)
