一种基于改进SIFT算法的遥感影像配准方法A registration method of RS image based on improved SIFT algorithm
雷小群,李芳芳,肖本林
摘要(Abstract)：
本文首先分析了传统SIFT算法在遥感影像配准应用中存在的不足,从特征点匹配精度上进行了改进。通过分析SIFT特征点匹配的主要误差来源,逐步消除可能误差,提取尽可能多且精准的匹配点对。接下来利用该匹配点对作为配准用控制点对,分别对不同时相、不同分辨率遥感影像进行仿射变换和小面元微分纠正配准。最后与传统SIFT算法进行比较,试验结果表明本文算法具有更高的匹配精度。
关键词(KeyWords)： SIFT算法;小面元微分纠正;影像配准
基金项目(Foundation): 湖北省自然科学基金资助项目(2008CDB388)
作者(Author): 雷小群,李芳芳,肖本林
DOI: 10.16251/j.cnki.1009-2307.2010.03.071
参考文献(References)：
- [1]Morave'c H.Visual Mapping by a Robot Rover[C]//Proc 6th Int Joint Conf Artificial Intell Tokyo,Japan,1979-08.
- [2]Forstner W A.Feature based Correspondence Algorithmfor Image Matching.Rovaniemi,Finland,Int Arch ofPhotogrammetry,1986;26-Ⅲ.
- [3]Hannah M J.Computer Matching of Areas in Stereo Im-agery[D]//Ph D dissertation,AIM239.Stanford Cal-ifornia,USA:Computer Science Department,StanfordUniversity,1974.
- [4]Luhmann T,Altrogge G.Interest Operator for imageMatching[C]//Intercommission Conference on FastProcessing of Photogrammetric Data,Interlaken,Switz-erland,1987-06.
- [5]D G Lowe.Object Recognition from Local Scale-Invari-ant Features[C]//International Conference on Com-puter Vision,Corfu,Greece,1999:1150-1157.
- [6]D G Lowe,Distinctive image features from scale-invari-ant keypoints[J].International Journal of ComputerVision,2004,60(2).
- [7]Mikolajczyk K,Schmid C.A Performance Evaluation ofLocal Descriptors[J].IEEE Trans.Pattern Analysisand Machine Intelligence,2005,27(10).
- [8]Brown M,Lowe D G.Recognizing Panoramas[C]//In Proceedings of the 9th International Conference onComputer Vision(ICCV03).Nice,October,2003.
- [9]Schafalitzky F,Zisserm an A.Multi-view Matching forUnordered Image Sets,or Howdo I Organize my HolidaySnaps-[C]//Proceedings of the 7th European Confer-ence on Computer Vision(ECCV02),2002.
- [10]M Fischler,R Bolles.Random Sample Consensus:AParadigm for Model Fitting With Applications to ImageAnalysis and Automated Cartography[C]//ACM,Graphics and Image Processing,1981.
- [11]李二森,等.SIFT特征匹配技术在自动相对定向中的应用[J].测绘科学,2008,33(5).