On the geometry of multivariate generalized Gaussian models. (English) Zbl 1255.68199

Verdoolaege, Geert; Scheunders, Paul

J. Math. Imaging Vis. 43, No. 3, 180-193 (2012).

Summary: This paper concerns the geometry of the zero-mean multivariate generalized Gaussian distribution (MGGD) and the calculation of geodesic distances on the MGGD manifold. The MGGD is a suitable distribution for the modeling of multivariate (color, multispectral, vector and tensor images, etc.) image wavelet statistics. Expressions are derived for the Fisher-Rao metric for the zero-mean MGGD model. A closed-form expression is obtained for the geodesic distance on the submanifolds characterized by a fixed MGGD shape parameter. Suitable approximate solutions to the geodesic equations are presented in the case of MGGDs with varying shape parameters. An application to image texture similarity measurement in the wavelet domain is briefly discussed, comparing the performance of the geodesic distance and the Kullback-Leibler divergence.

MSC:

68U05	Computer graphics; computational geometry (digital and algorithmic aspects)
62H35	Image analysis in multivariate analysis

Keywords:

geodesic distance; multivariate generalized Gaussian distribution; information geometry; multicomponent texture discrimination

References:

[1]	Amari, S., Nagaoka, H.: Methods of Information Geometry. Transactions of Mathematical Monographs, vol. 191. American Mathematical Society, New York (2000) · Zbl 0960.62005
[2]	Atkinson, C., Mitchell, A.: Rao’s distance measure. Sankhya, Ser. A 48, 345–365 (1981) · Zbl 0534.62012
[3]	Berkane, M., Oden, K., Bentler, P.: Geodesic estimation in elliptical distributions. J. Multivar. Anal. 63(1), 35–46 (1997) · Zbl 0897.62052
[4]	Berman, A., Shapiro, L.: A flexible image database system for content-based retrieval. Comput. Vis. Image Underst. 75(1–2), 175–195 (1999) · Zbl 05467776
[5]	Burbea, J.: Informative geometry of probability spaces. Expo. Math. 4, 347–378 (1986) · Zbl 0604.62006
[6]	Burbea, J., Rao, C.: Entropy differential metric, distance and divergence measures in probability spaces: a unified approach. J. Multivar. Anal. 12(4), 575–596 (1982) · Zbl 0526.60015
[7]	Burkhard, W., Keller, R.: Some approaches to best-match file searching. Commun. ACM 16(4), 230–236 (1973) · Zbl 0269.68062
[8]	Calvo, M., Oller, J.: A distance between elliptical distributions based in an embedding into the Siegel group. J. Comput. Appl. Math. 145(2), 319–334 (2002) · Zbl 1021.62041
[9]	Caselles, V., Kimmel, R., Sapiro, G.: Geodesic active contours. Int. J. Comput. Vis. 22(1), 61–79 (1997) · Zbl 0894.68131
[10]	Castano-Moraga, C., Lenglet, C., Deriche, R., Ruiz-Alzola, J.: A Riemannian approach to anisotropic filtering of tensor fields. Signal Process. 87(2), 263–276 (2007) · Zbl 1186.94077
[11]	Čenkov, N.: Statistical Decision Rules and Optimal Inference. Translations of Mathematical Monographs, vol. 53. American Mathematical Society, Providence (1982)
[12]	Cho, D., Bui, T.: Multivariate statistical modeling for image denoising using wavelet transforms. Signal Process. 20(1), 77–89 (2005)
[13]	Cramér, H.: A contribution to the theory of statistical estimation. Skand. Aktuarietidskr. 29, 85–94 (1946) · Zbl 0060.30513
[14]	Devijver, P., Kittler, J.: Pattern Recognition: A Statistical Approach. Prentice-Hall, London (1982) · Zbl 0542.68071
[15]	Do, M., Vetterli, M.: Wavelet-based texture retrieval using generalized Gaussian density and Kullback-Leibler distance. IEEE Trans. Image Process. 11(2), 146–158 (2002) · Zbl 05452760
[16]	do Carmo, M., Flaherty, F.: Riemannian Geometry. Birkhäuser, Boston (1992)
[17]	Fang, K.T., Zhang, Y.T.: Generalized Multivariate Analysis. Springer, Berlin (1990a) · Zbl 0724.62054
[18]	Fang, K.T., Zhang, Y.T.: Generalized multivariate analysis. In: Generalized Multivariate Analysis. Springer, Berlin (1990b), Sect. 1.4.2 · Zbl 0724.62054
[19]	Fang, K.T., Kotz, S., Ng, K.W.: Symmetric Multivariate and Related Distributions. Monographs on Statistics and Applied Probability, vol. 36. Chapman and Hall, New York (1990) · Zbl 0699.62048
[20]	Gómez, E., Gómez-Villegas, M., Marín, J.: A multivariate generalization of the power exponential family of distributions. Commun. Stat., Theory Methods 27(3), 589–600 (1998) · Zbl 0895.62053
[21]	Gong, M., Jiao, L., Bo, L., Wang, L., Zhang, X.: Image texture classification using a manifold-distance-based evolutionary clustering method. Opt. Eng. 47(7), 077,201 (2008)
[22]	Hua, J., Lai, Z., Dong, M., Qin, H., Gu, X.: Geodesic distance-weighted shape vector image diffusion. IEEE Trans. Vis. Comput. Graph. 14(6), 1643–1650 (2008)
[23]	Huot, E., Yahia, H., Cohen, I., Herlin, I.: Matching structures by computing minimal paths on a manifold. J. Vis. Commun. Image Represent. 13(1–2), 302–312 (2002)
[24]	Ikonen, L.: Priority pixel queue algorithm for geodesic distance transforms. Image Vis. Comput. 25(10), 1520–1529 (2007)
[25]	James, A.: The variance information manifold and the functions on it. In: Krishnaiah, P. (ed.) Multivariate Analysis III, pp. 157–169. Academic Press, New York (1973)
[26]	Jeffreys, H., Jeffreys, B.S.: Methods of Mathematical Physics, 3rd edn. Cambridge University Press, Cambridge (2000), Sect. 15.08
[27]	Kass, R., Vos, P.: Geometrical Foundations of Asymptotic inference. Wiley Series in Probability and Statistics. Wiley-Interscience, New York (1997) · Zbl 0880.62005
[28]	Kotz, S.: Multivariate distributions at a cross road. In: Statistical Distributions in Scientific Work, pp. 247–270. Reidel, Dordrecht (1968)
[29]	Kullback, S.: Information Theory and Statistics. Dover, New York (1968) · Zbl 0274.62036
[30]	Lenglet, C., Rousson, M., Deriche, R.: DTI segmentation by statistical surface evolution. IEEE Trans. Med. Imaging 25(6), 685–700 (2006a)
[31]	Lenglet, C., Rousson, M., Deriche, R., Faugeras, O.: Statistics on the manifold of multivariate normal distributions: theory and application to diffusion tensor MRI processing. J. Math. Imaging Vis. 25(3), 423–444 (2006b) · Zbl 1478.62387
[32]	Li, L., Xu, C., Tang, W., Zhong, C.: 3D face recognition by constructing deformation invariant image. Pattern Recogn. Lett. 29(10), 1596–1602 (2008)
[33]	Mallat, S.: A theory for multiresolution signal decomposition: the wavelet representation. IEEE Trans. Pattern Anal. Mach. Intell. 11(7), 674–692 (1989) · Zbl 0709.94650
[34]	Mardia, K., Kent, J., Bibby, J.: Multivariate Analysis. Academic Press, London (1982)
[35]	Maybank, S.: Application of the Fisher-Rao metric to ellipse detection. Int. J. Comput. Vis. 72(3), 287–307 (2007) · Zbl 1477.68397
[36]	Meng, D., Leung, Y., Xu, Z., Fung, T., Zhang, Q.: Improving geodesic distance estimation based on locally linear assumption. Pattern Recogn. Lett. 29(7), 862–870 (2008)
[37]	MIT Vision and Modeling Group (2010) Vision texture. Online at http://vismod.media.mit.edu/vismod/imagery/VisionTexture/
[38]	Mitchell, A.: The information matrix, skewness tensor and {\(\alpha\)}-connections for the general multivariate elliptic distribution. Ann. Inst. Stat. Math. 41(2), 289–304 (1989) · Zbl 0691.62049
[39]	Muirhead, R.: Aspects of Multivariate Statistical Theory, 2nd edn. Wiley Series in Probability and Statistics. Wiley-Interscience, New York (2005), Sect. 1.5 · Zbl 0556.62028
[40]	Murray, M., Rice, J.: Differential Geometry and Statistics. Monographs on Statistics and Applied Probability, vol. 48. Chapman and Hall, New York (1993) · Zbl 0804.53001
[41]	O’Neill, B.: Elementary Differential Geometry, 2nd edn. Academic Press, New York (1982)
[42]	Pastore, J., Moler, E., Ballarin, V.: Segmentation of brain magnetic resonance images through morphological operators and geodesic distance. Digit. Signal Process. 15(2), 153–160 (2005)
[43]	Pennec, X., Fillard, P., Ayache, N.: A Riemannian framework for tensor computing. Int. J. Comput. Vis. 66(1), 41–66 (2006) · Zbl 1287.53031
[44]	Rao, C.: Information and accuracy attainable in the estimation of statistical parameters. Bull. Calcutta Math. Soc. 37, 81–89 (1945) · Zbl 0063.06420
[45]	Skovgaard, L.: A Riemannian geometry of the multivariate normal model. Tech. rep. 81/3, Statistical Research Unit, Danish Medical Research Council, Danish Social Science Research Council (1981) · Zbl 0477.62009
[46]	Skovgaard, L.: A Riemannian geometry of the multivariate normal model. Scand. J. Stat. 11(4), 211–223 (1984) · Zbl 0579.62033
[47]	Tenenbaum, J., de Silva, V., Langford, J.: A global geometric framework for nonlinear dimensionality reduction. Science 290(5500), 2319–2323 (2000)
[48]	The Mathworks (2008) Natick, MA, www.mathworks.com
[49]	Theodoridis, S., Koutrombas, K.: Pattern Recognition, 2nd edn. Academic Press, London (2003), Sect. B.2
[50]	Twining, C., Marsland, S.: Constructing an atlas for the diffeomorphism group of a compact manifold with boundary, with application to the analysis of image registrations. J. Comput. Appl. Math. 222(2), 411–428 (2008) · Zbl 1152.58004
[51]	Verdoolaege, G., Scheunders, P.: Geodesics on the manifold of multivariate generalized Gaussian distributions with an application to multicomponent texture discrimination. Int. J. Comput. Vis. (2011) · Zbl 1235.68316
[52]	Verdoolaege, G., De Backer, S., Scheunders, P.: Multiscale colour texture retrieval using the geodesic distance between multivariate generalized Gaussian models. In: Proceedings of the 15th IEEE International Conference on Image Processing, pp. 169–172 (2008)
[53]	Yong, Q., Jie, Y.: Modified kernel functions by geodesic distance. EURASIP J. Appl. Signal Process. 16, 2515–2521 (2004) · Zbl 1107.68471