A Novel Iterative SIFT Points Matching Method

This work focuses on feature point matching problem based on local descriptor scale-invariant feature transform (SIFT). Feature point matching is a fundamental problem in image analysis and computer vision, and SIFT is one of the most typical local descriptors. Essentially, the available methods of SIFT matching treat the feature points equally in matching, which results in very sparse matches. To increase the density of matched SIFT points, we propose a scheme to improve the performance of SIFT matching by mining additional information from the topological relationships between feature points and from the matching procedures. The fundamental idea is that the distinctiveness are different from point to point, some are easy to be matched correctly others are relatively unreliable to be matched directly; and we can discover some supplementary information from matched points to reduce ambiguity of the less distinctive points and improve their matching accuracy. Experiments verify the performance of the proposed method.

[1]  Shuicheng Yan,et al.  Correspondence Propagation with Weak Priors , 2009, IEEE Transactions on Image Processing.

[2]  Chao Wang,et al.  Improved SIFT-Features Matching for Object Recognition , 2008, BCS Int. Acad. Conf..

[3]  Milan Sonka,et al.  Image Processing, Analysis and Machine Vision , 1993, Springer US.

[4]  Jiri Matas,et al.  Improving Descriptors for Fast Tree Matching by Optimal Linear Projection , 2007, 2007 IEEE 11th International Conference on Computer Vision.

[5]  Desire Sidibé,et al.  Fast and robust image matching using contextual information and relaxation , 2007, VISAPP.

[6]  Xiao-Ming Li,et al.  Rejecting Outliers Based on Correspondence Manifold , 2009 .

[7]  Alexandre Bernardino,et al.  Improving the SIFT descriptor with smooth derivative filters , 2009, Pattern Recognit. Lett..

[8]  Arturo Espinosa-Romero,et al.  A robust Graph Transformation Matching for non-rigid registration , 2009, Image Vis. Comput..

[9]  Michal Perdoch,et al.  Efficient sequential correspondence selection by cosegmentation , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[10]  Martial Hebert,et al.  A spectral technique for correspondence problems using pairwise constraints , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.

[11]  Francesca Odone,et al.  SVD-matching using SIFT features , 2006, Graph. Model..

[12]  David G. Lowe,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004, International Journal of Computer Vision.

[13]  Zhanyi Hu,et al.  Rejecting Mismatches by Correspondence Function , 2010, International Journal of Computer Vision.

[14]  Michael Werman,et al.  A Linear Time Histogram Metric for Improved SIFT Matching , 2008, ECCV.

[15]  Michael Werman,et al.  Fast and robust Earth Mover's Distances , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[16]  Zhanyi Hu,et al.  This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. IEEE TRANSACTION ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 1 Rotationally Invariant Descript , 2011 .

[17]  David G. Lowe,et al.  Object recognition from local scale-invariant features , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[18]  Gertjan J. Burghouts,et al.  Performance evaluation of local colour invariants , 2009, Comput. Vis. Image Underst..

[19]  Mao-Yong Cao,et al.  Rejecting Outliers Based on Correspondence Manifold: Rejecting Outliers Based on Correspondence Manifold , 2009 .

[20]  Philippe Montesinos,et al.  Matching Local Invariant Features with Contextual Information: An Experimental Evaluation , 2008 .

[21]  Cordelia Schmid,et al.  A Performance Evaluation of Local Descriptors , 2005, IEEE Trans. Pattern Anal. Mach. Intell..