False Detection of Relative Orientation of Multi-view Images Based on Belief Propagation

doi:10.11947/j.AGCS.2015.20140255

Abstract

Abstract: Ambiguous visual structures or repetitive textures of different scenarios would result in a large number of mismatching points in image matching stage, and false positive relative orientation is introduced which is locally consistent.In this paper, the probabilistic inference method is introduced and an improved detection algorithm of false positive relative orientations is presented. A Bayes network based on relative orientations between images is constructed using closed cycle constraints and priori probability models in Bayes network is derived and the maximum posterior probability in Bayesian network is solved using belief propagation algorithm. Experimental results show that using the proposed global consistency constraints method can effectively detect the false positive relative orientations and improve outcomes of scene reconstruction with high computational efficiency.

Key words: relative orientation, cycle constraint, maximum spanning tree, Bayes network, belief propagation

CLC Number:

P237

LU Jun, ZHANG Baoming, GUO Haitao, ZHANG Hongwei. False Detection of Relative Orientation of Multi-view Images Based on Belief Propagation[J]. Acta Geodaetica et Cartographica Sinica, 2015, 44(4): 422-430.

References

[1] SNAVELY N, SEITZ S M, SZELISKI R. Modeling the World from Internet Photo Collections[J]. International Journal of Computer Vision, 2008, 80(2): 189-210.
[2] SNAVELY N, SEITZ S M, SZELISKI R. Skeletal Graphs for Efficient Structure from Motion[C]//IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, AK: IEEE,2008:1-8.
[3] AGARWAL S, FURUKAWA Y, SNAVELY N, et al. Building Rome in a Day[J]. Communications of the ACM, 2011, 54(10): 105-112.
[4] KAHL F. Multiple View Geometry and the L_∞-norm[C]//Tenth IEEE International Conference on Computer Vision. Beijing: IEEE, 2005, 2: 1002-1009.
[5] SINHA S N, STEEDLY D, SZELISKI R. A Multi-stage Linear Approach to Structure from Motion[M]//KUTULAKOS K N ed. Trends and Topics in Computer Vision. Berlin:Springer, 2012: 267-281.
[6] ARIE-NACHIMSON M, KOVALSKY S Z, KEMELMACHER-SHLIZERMAN I, et al. Global Motion Estimation from Point Matches[C]//2012 Second International Conference on 3D Imaging, Modeling, Processing, Visualization and Transmission (3DIMPVT). Zurich: IEEE, 2012: 81-88.
[7] MOULON P, MONASSE P, MARLET R. Global Fusion of Relative Motions for Robust, Accurate and Scalable Structure from Motion[C]//2013 IEEE International Conference on Computer Vision (ICCV). Sydney, NSW:IEEE, 2013: 3248-3255.
[8] MARTINEC D, PAJDLA T. Robust Rotation and Translation Estimation in Multiview Reconstruction[C]//IEEE Conference on Computer Vision and Pattern Recognition. Minneapolis, MN:IEEE, 2007: 1-8.
[9] WANG Jingxue, ZHU Qing, WANG Weixi. A Dense Matching Algorithm of Multi-view Image Based on the Integrated Multiple Matching Primitives [J]. Acta Geodaetica et Cartographica Sinica, 2013, 42(5): 691-698. (王竞雪, 朱庆, 王伟玺. 多匹配基元集成的多视影像密集匹配方法[J]. 测绘学报, 2013, 42(5): 691-698.)
[10] YUAN Xiuxiao, MING Yang. A Novel Method of Multi-image Matching Using Image and Space Synthesis Information [J].Acta Geodaetica et Cartographica Sinica, 2009, 38(3): 216-222. (袁修孝, 明洋. 一种综合利用像方和物方信息的多影像匹配方法[J]. 测绘学报, 2009, 38(3): 216-222.)
[11] JIA Fengman, KANG Zhizhong, YU Peng. A SIFT and Bayes Sampling Consensus Method for Image Matching [J]. Acta Geodaetica et Cartographica Sinica, 2013, 42(6): 877-883. (贾丰蔓, 康志忠, 于鹏. 影像同名点匹配的 SIFT 算法与贝叶斯抽样一致性检验[J]. 测绘学报, 2013, 42(6): 877-883.)
[12] GOVINDU V M. Robustness in Motion Averaging[M]// NARAYANAN P J, NAYAR S K, SHUM H Y.Computer Vision:ACCV 2006. Berlin:Springer, 2006: 457-466.
[13] OLSSON C, ENQVIST O. Stable Structure from Motion for Unordered Image Collections[M]//HEYDEN A, KAHL F.Image Analysis. Berlin:Springer, 2011: 524-535.
[14] ENQVIST O, KAHL F, OLSSON C. Non-sequential Structure from Motion[C]//2011 IEEE International Conference on Computer Vision Workshops (ICCV Workshops). Barcelona:IEEE, 2011: 264-271.
[15] LOWE D G. Distinctive Image Features from Scale-invariant Key Points[J]. International Journal of Computer Vision, 2004, 60(2): 91-110.
[16] NISTÉR D.An Efficient Solution to the Five-point Relative Pose Problem[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2004, 26(6): 756-770.
[17] KRUSKAL J B. On the Shortest Spanning Subtree of A Graph and the Traveling Salesman Problem[J]. Proceedings of the American Mathematical Society, 1956, 7(1): 48-50.
[18] ZACH C, KLOPSCHITZ M, POLLEFEYS M. Disambiguating Visual Relations Using Loop Constraints[C]//2010 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). San Francisco, CA:IEEE, 2010: 1426-1433.
[19] PEARL J. Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference[M]. San Francisco:Morgan Kaufmann, 1988.
[20] MACKAY D J C, NEAL R M. Good Codes Based on Very Sparse Matrices[M]//BOYD C ed. Cryptography and Coding. Berlin:Springer, 1995: 100-111.
[21] KSCHISCHANG F R, FREY B J. Iterative Decoding of Compound Codes by Probability Propagation in Graphical Models[J]. IEEE Journal on Selected Areas in Communications, 1998, 16(2): 219-230.
[22] MCELIECE R J, MACKAY D J C, CHENG J F. Turbo Decoding as an Instance of Pearl's “Belief Propagation” Algorithm[J]. IEEE Journal on Selected Areas in Communications, 1998, 16(2): 140-152.
[23] KSCHISCHANG F R, FREY B J, LOELIGER H A. Factor Graphs and the Sum-product Algorithm[J]. IEEE Transactions on Information Theory, 2001, 47(2): 498-519.
[24] MURPHY K P, WEISS Y, JORDAN M I. Loopy Belief Propagation for Approximate Inference: An Empirical Study[C]//Proceedings of the Fifteenth Conference on Uncertainty in Artificial Intelligence. San Francisco:Morgan Kaufmann Publishers Inc., 1999: 467-475.