Abstract
A classical solution for matching two image patches is to use the cross-correlation coefficient. This works well if there is a lot of structure within the patches, but not so well if the patches are close to uniform. This means that some patches are matched with more confidence than others. By estimating this uncertainty, more weight can be put on the confident matches than those that are more uncertain. To enable this two distribution functions for two different cases are used: (i) the correlation between two patches showing the same object but with different lighting conditions and different noise realisations and (ii) the correlation between two unrelated patches.
Using these two distributions the patch matching problem is, in this paper, formulated as a binary classification problem. The probability of two patches matching is derived. The model depends on the signal to noise ratio. The noise level is reasonably invariant over time, while the signal level, represented by the amount of structure in the patch or its spatial variance, has to be measured for every frame.
A common application where this is useful is feature point matching between different images. Another application is background/foreground segmentation. This paper will concentrate on the latter application. It is shown how the theory can be used to implement a very fast background/foreground segmentation algorithm by transforming the calculations to the DCT-domain and processing a motion-JPEG stream without uncompressing it. This allows the algorithm to be embedded on a 150 MHz ARM based network camera. It is also suggested to use recursive quantile estimation to estimate the background model. This gives very accurate background models even if there is a lot of foreground present during the initialisation of the model.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Abramowitz, M., Stegun, I.A.: Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables. Dover, New York (1972), 8. pr. with corr. edn
Ardö, H.: Multi-target Tracking Using On-line Viterbi Optimisation and Stochastic Modelling. Centre for Mathematical Sciences LTH, Lund University, Sweden (2009)
Ardö, H., Ăström, K.: Bayesian formulation of image patch matching using cross-correlation. In: Third ACM/IEEE International Conference on Distributed Smart Cameras (2009)
Brunelli, R., Poggio, T.: Face recognition: features versus templates. IEEE Trans. Pattern Anal. Mach. Intell. 15(10), 1042–1052 (1993)
Duda, R.O., Hart, P.E.: Pattern Classification and Scene Analysis. Wiley–Interscience, New York (1973)
Friedman, N., Russell, S.: Image segmentation in video sequences: a probabilistic approach. In: Thirteenth Conference on Uncertainty in Artificial Intelligence, pp. 175–181 (1997). doi:10.1.1.38.7356. URL http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.38.7356
Gordon, G., Darrell, T., Harville, M., Woodfill, J.: Background estimation and removal based on range and color. In: IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 1999, vol. 2, p. 464 (1999). doi:10.1109/CVPR.1999.784721. URL http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=784721
Hotelling, H.: New light on the correlation coefficient and its transforms. J. R. Stat. Soc., Ser. B 15(2), 193–232 (1953)
Hu, W., Gong, H., Zhu, S.C., Wang, Y.: An integrated background model for video surveillance based on primal sketch and 3d scene geometry. In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2008, pp. 1–8 (2008). doi:10.1109/CVPR.2008.4587541. URL http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4587541
Johnson, N.L.: Continuous Univariate Distributions, vol. 2. Wiley, New York (1995)
Kaneko, S., Murase, I., Igarashi, S.: Robust image registration by increment sign correlation. Pattern Recognit. 35, 2223–2234 (2002)
Kohli, P., Torr, P.: Efficiently solving dynamic Markov random fields using graph cuts. In: ICCV 2005, vol. 2, pp. 922–929 (2005)
Mittal, A., Paragios, N.: Motion-based background subtraction using adaptive kernel density estimation. In: CVPR, vol. 02, pp. 302–309 (2004). URL http://doi.ieeecomputersociety.org/10.1109/CVPR.2004.164
Möller, E., Grieszbach, G., Schack, B., Witte, H., Maurizio, P.: Statistical properties and control algorithms of recursive quantile estimators. Biom. J. 42(6), 729–746 (2000)
Noriega, P., Bernier, O.: Real time illumination invariant background subtraction using local kernel histograms. In: Proc. British Machine Vision Conference, p. III:979 (2006)
Spanne, S.: Konkret Analys. KFS AB, Solvegatan 22, Lund, Sweden (1997)
Stauffer, C.: Adaptive background mixture models for real-time tracking. In: Proc. Conf. Computer Vision and Pattern Recognition, pp. 246–252 (1999)
Sullivan, J., Blake, M., Isard, M., MacCormick, J.: Bayesian object localisation in images. Int. J. Comput. Vis. 44(2), 111–135 (2001)
Toyama, K., Krumm, J., Brumitt, B., Meyers, B.: Wallflower: principles and practice of background maintenance. In: Int. Conf. on Computer Vision, pp. 255–261 (1999)
Magnus, W.F., Bateman, H., Erdaelyi, A.: Higher Transcendental Functions, vol. II. McGraw–Hill, New York (1953)
Wayne, P., Johann, P., Schoonees, A.: Understanding background mixture models for foreground segmentation. In: Proceedings Image and Vision Computing (2002). URL http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.9.5222
Wren, C.R., Azarbayejani, A., Darrell, T., Pentland, A.P.: Pfinder: real-time tracking of the human body. IEEE Trans. Pattern Anal. Mach. Intell. 19(7), 780–785 (1997)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Ardö, H., Ågström, K. Bayesian Formulation of Image Patch Matching Using Cross-correlation. J Math Imaging Vis 43, 72–87 (2012). https://doi.org/10.1007/s10851-011-0287-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10851-011-0287-x