Fast image matching algorithm based on affine invariants

Abstract

Feature-based image matching algorithms play an indispensable role in automatic target recognition (ATR). In this work, a fast image matching algorithm (FIMA) is proposed which utilizes the geometry feature of extended centroid (EC) to build affine invariants. Based on affine invariants of the length ratio of two parallel line segments, FIMA overcomes the invalidation problem of the state-of-the-art algorithms based on affine geometry features, and increases the feature diversity of different targets, thus reducing misjudgment rate during recognizing targets. However, it is found that FIMA suffers from the parallelogram contour problem and the coincidence invalidation. An advanced FIMA is designed to cope with these problems. Experiments prove that the proposed algorithms have better robustness for Gaussian noise, gray-scale change, contrast change, illumination and small three-dimensional rotation. Compared with the latest fast image matching algorithms based on geometry features, FIMA reaches the speedup of approximate 1.75 times. Thus, FIMA would be more suitable for actual ATR applications.

Keywords

affine invariants / image matching / extended centroid / robustness / performance

References

[1] ZitovaB, FlusserJ. Image registration methods: A survey [J]. Image and Vision Computing, 2003, 21(11): 977-1000

[2] HuM K. Visual pattern recognition by moment invariants [J]. IRE Transactions on Information Theory, 1962, 8(2): 179-187

[3] LeiL, ChenT, LiZ-y, SuYi. A New approach to image invariant extraction under global affine transformation [J]. Journal of National University of Defense Technology, 2008, 4: 64-70

[4] LisinD A, MattarM A, BlaschkoM B. Combining local and global image features for object class recognition [C]. IEEE Computer Society Conference on Computer Vision and Pattern Recognition-Workshops, 2005, Washington DC, IEEE: 47-47

[5] LoweD G. Distinctive image features from scale-invariant keypoints [J]. International Journal of Computer Vision, 2004, 60(2): 91-110

[6] BayH, TuytelaarsT, GoolLSurf: Speeded up robust features [M]. Computer Vision-ECCV 2006, 2006, Graz, Austria, Springer Berlin Heidelberg: 404-417

[7] MatasJ, ChumO, MartinU, PajdlaT. Robust wide baseline stereo from maximally stable external regions [J]. Image and Vision Computing, 2004, 22(10): 761-767

[8] AmiriM, RabieeH R. RASIM: A novel rotation and scale invariant matching of local image interest points [J]. IEEE Transactions on Image Processing, 2011, 20(12): 3580-3591

[9] RahtuE, SaloM, HeikkilaJ. Affine invariant pattern recognition using multi scale auto convolution [J]. IEEE Transactions on PAMI, 2005, 27(6): 908-918

[10] ChenT, SuY, JiangY-m, TangT, YuW-xian. Affine invariant feature extraction based on affine geometry [J]. Journal of Image and Graphics, 2007, 12(9): 1633-1641

[11] GaoF, WenG-jian. A new method for affine invariants extraction based on affine geometry [J]. Journal of Image and Graphics, 2011, 16(3): 389-397

[12] ChristoD, PrinceJ L, BryanR N. Image registration based on boundary mapping [J]. IEEE Transactions on Medical Imaging, 1996, 15(1): 112-115

[13] YangZ, CohenF S. Image registration and object recognition using affine invariants and convex hulls [J]. IEEE Transactions on Image Processing, 1999, 8(7): 934-946

[14] BianX-b, ZhuQ-xin. Analysis of moment invariants stability to gray-level adjustment [C]. Proceedings of ICSSM’05, 2005, Washington DC, USA, IEEE Computer Society: 1145-1148

[15] NeneS A, NayarS K, MuraseHColumbia object image library (COIL-100) [R]. Technical Report CUCS-005-96, 1996, New York, Columbia University

[16] QuZ-g, WangP, GaoY-hui. Randomized SUSAN edge detector [J]. Optical Engineering, 2011, 50(11): 110502-110502-3

[17] YuY-n, HuangK-q, ChenW, TanT-niu. A novel algorithm for view and illumination invariant image matching [J]. IEEE Transactions on Image Processing, 2012, 21(1): 229-240

[18] MorelJ M, YuG. ASIFT: A new framework for fully affine invariant image comparison [J]. SIAM J Imag Sci, 2009, 2(2): 438-469

[19] HarderM, PolaniD, NehanivC L. Think globally, sense locally: From local information to global features [C]. 2011 IEEE Symposium on Artificial Life, 2011, Paris, IEEE: 70-77

[20] KuruzB J. Optimal color quantization for color displays [C]. IEEE Conference on Computer Vision and Pattern Recognition. Washington DC, USA, 1983217-224

[21] XiangZ-gan. Color image quantization by minimizing the maximum intercluster distance [J]. ACM Transactions on Graphics, 1997, 16(3): 260-276

[22] ZhangY, LuK, GaoY-h, WangMo. NEQR: A novel enhanced quantum representation of digital images [J]. Quantum Information Processing, 2013, 12(8): 2833-2860

[23] ZhangY, LuK, GaoY-h, XuKai. A novel quantum representation for log-polar images [J]. Quantum Information Processing, 2013, 12(9): 3103-3126