×

Geodesics on the manifold of multivariate generalized Gaussian distributions with an application to multicomponent texture discrimination. (English) Zbl 1235.68316

Summary: We consider the Rao geodesic distance (GD) based on the Fisher information as a similarity measure on the manifold of zero-mean multivariate generalized Gaussian distributions (MGGD). The MGGD is shown to be an adequate model for the heavy-tailed wavelet statistics in multicomponent images, such as color or multispectral images. We discuss the estimation of MGGD parameters using various methods. We apply the GD between MGGDs to color texture discrimination in several classification experiments, taking into account the correlation structure between the spectral bands in the wavelet domain. We compare the performance, both in terms of texture discrimination capability and computational load, of the GD and the Kullback-Leibler divergence (KLD). Likewise, both uni- and multivariate generalized Gaussian models are evaluated, characterized by a fixed or a variable shape parameter. The modeling of the interband correlation significantly improves classification efficiency, while the GD is shown to consistently outperform the KLD as a similarity measure.

MSC:

68U10 Computing methodologies for image processing
62H35 Image analysis in multivariate analysis
PDFBibTeX XMLCite
Full Text: DOI Link

References:

[1] Abramowitz, M., & Stegun, I. (1965). Handbook of mathematical functions. New York: Dover. · Zbl 0171.38503
[2] Amari, S., & Nagaoka, H. (2000). Methods of information geometry. Transactions of mathematical monographs: Vol. 191. New York: Am. Math. Soc. · Zbl 0960.62005
[3] Atkinson, C., & Mitchell, A. (1981). Rao’s distance measure. Sankhya. The Indian Journal of Statistics, 48, 345–365. · Zbl 0534.62012
[4] Benazza-Benyahia, A., & Pesquet, J. C. (2005). Building robust wavelet estimators for multicomponent images using Steins’ principle. IEEE Transactions on Image Processing, 14(11), 1814–1830. · Zbl 05452766 · doi:10.1109/TIP.2005.857247
[5] Berkane, M., Oden, K., & Bentler, P. (1997). Geodesic estimation in elliptical distributions. Journal of Multivariate Analysis, 63(1), 35–46. · Zbl 0897.62052 · doi:10.1006/jmva.1997.1690
[6] Berman, A., & Shapiro, L. (1999). A flexible image database system for content-based retrieval. Computer Vision and Image Understanding, 75(1–2), 175–195. · Zbl 05467776 · doi:10.1006/cviu.1999.0772
[7] Bonet, J., & Viola, P. (1998). Texture recognition using a nonparametric multi-scale statistical model. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 641–647). Santa Barbara, CA.
[8] Boubchir, L., & Fadili, J. (2005). Multivariate statistical modeling of images with the curvelet transform. In Proceedings of the 8th international symposium on signal processing and its applications (pp. 747–750).
[9] Burbea, J. (1986). Informative geometry of probability spaces. Expositiones Mathematicae, 4, 347–378. · Zbl 0604.62006
[10] Burbea, J., & Rao, C. (1982). Entropy differential metric, distance and divergence measures in probability spaces: a unified approach. Journal of Multivariate Analysis, 12, 575–596. · Zbl 0526.60015 · doi:10.1016/0047-259X(82)90065-3
[11] Burkhard, W., & Keller, R. (1973). Some approaches to best-match file searching. Communications of the ACM, 16(4), 230–236. · Zbl 0269.68062 · doi:10.1145/362003.362025
[12] Calvo, M., & Oller, J. (2002). A distance between elliptical distributions based in an embedding into the Siegel group. Journal of Computational and Applied Mathematics, 145(2), 319–334. · Zbl 1021.62041 · doi:10.1016/S0377-0427(01)00584-2
[13] Castano-Moraga, C., Lenglet, C., Deriche, R., & Ruiz-Alzola, J. (2007). A Riemannian approach to anisotropic filtering of tensor fields. Signal Processing, 87(2), 263–276. · Zbl 1186.94077 · doi:10.1016/j.sigpro.2006.02.049
[14] Čenkov, N. (1982). Statistical decision rules and optimal inference. Translations of mathematical monographs: Vol. 53. Providence: Am. Math. Soc.
[15] Chang, C. I. (2007). Hyperspectral data exploitation: theory and applications. New York: Wiley-Interscience.
[16] Cho, D., & Bui, T. (2005). Multivariate statistical modeling for image denoising using wavelet transforms. Signal Processing, 20(1), 77–89.
[17] Cramér, H. (1946). A contribution to the theory of statistical estimation. Skandinavisk Aktuarietidskrift, 29, 85–94. · Zbl 0060.30513
[18] Datta, R., Joshi, D., Li, J., & Wang, J. (2008). Image retrieval: Ideas, influences, and trends of the new age. ACM Computing Surveys, 40(2).
[19] Daugman, J. (1980). Two-dimensional spectral analysis of cortical receptive field profile. Vision Research, 20(10), 847–856. · doi:10.1016/0042-6989(80)90065-6
[20] De Backer, S., Pizurica, A., Huysman, B., Philips, W., & Scheunders, P. (2008). Denoising of multispectral images using wavelet least-squares estimators. Image and Vision Computing, 26(7), 1038–1051. · doi:10.1016/j.imavis.2007.11.003
[21] do Carmo, M., & Flaherty, F. (1992). Riemannian geometry. Boston: Birkhäuser.
[22] Do, M., & Vetterli, M. (2002). Wavelet-based texture retrieval using generalized Gaussian density and Kullback-Leibler distance. IEEE Transactions on Image Processing, 11(2), 146–158. · Zbl 05452760 · doi:10.1109/83.982822
[23] Fang, K. T., & Zhang, Y. T. (1990). Generalized multivariate analysis. Berlin: Springer. · Zbl 0724.62054
[24] Fang, K. T., Kotz, S., & Ng, K. W. (1990). Symmetric multivariate and related distributions. Monographs on statistics and applied probability: Vol. 36. New York: Chapman & Hall. · Zbl 0699.62048
[25] Gómez, E., Gómez-Villegas, M., & Marín, J. (1998). A multivariate generalization of the power exponential family of distributions. Communications in Statistics. Theory and Methods, 27(3), 589–600. · Zbl 0895.62053 · doi:10.1080/03610929808832115
[26] Imagefter (2010). Online at http://www.imageafter.com/ .
[27] James, A. (1973). The variance information manifold and the functions on it. In P. Krishnaiah (Ed.), Multivariate analysis III (pp. 157–169). New York: Academic Press.
[28] Jensen, S. T. (1976). Private communication.
[29] Kass, R., & Vos, P. (1997). Geometrical foundations of asymptotic inference. Wiley series in probability and statistics. New York: Wiley-Interscience. · Zbl 0880.62005
[30] Kullback, S. (1968). Information theory and statistics. New York: Dover. · Zbl 0274.62036
[31] Lehmann, E., & Casella, G. (2003). Theory of point estimation (2nd ed.). Springer texts in statistics. New York: Springer.
[32] Lenglet, C., Rousson, M., & Deriche, R. (2006a). DTI segmentation by statistical surface evolution. IEEE Transactions on Medical Imaging, 25(6), 685–700. · doi:10.1109/TMI.2006.873299
[33] Lenglet, C., Rousson, M., Deriche, R., & Faugeras, O. (2006b). Statistics on the manifold of multivariate normal distributions: Theory and application to diffusion tensor MRI processing. Journal of Mathematical Imaging and Vision, 25(3), 423–444. · Zbl 1478.62387 · doi:10.1007/s10851-006-6897-z
[34] Mallat, S. (1989). A theory for multiresolution signal decomposition: The wavelet representation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 11(7), 674–692. · Zbl 0709.94650 · doi:10.1109/34.192463
[35] Mallat, S. (1999). A wavelet tour of signal processing (2nd ed.). New York: Academic Press. · Zbl 0998.94510
[36] Manjunath, B., & Ma, W. (1996). Texture features for browsing and retrieval of image data. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8), 837–842. · Zbl 05112135 · doi:10.1109/34.531803
[37] Manjunath, B., Ohm, J. R., Vasudevan, V., & Yamada, A. (2001). Color and texture descriptors. IEEE Transactions on Circuits and Systems for Video Technology, 11(6), 703–715. · Zbl 05451252 · doi:10.1109/76.927424
[38] Mardia, K., Kent, J., & Bibby, J. (1982). Multivariate analysis. London: Academic Press. · Zbl 0432.62029
[39] Marsaglia, G. (1972). Choosing a point from the surface of a sphere. Annals of Mathematical Statistics, 43(2), 645–646. · Zbl 0248.65008 · doi:10.1214/aoms/1177692644
[40] Mathiassen, J., Skavhaug, A., & Bo, K. (2002). Texture similarity measure using Kullback-Leibler divergence between gamma distributions. In Proceedings of the European conference on computer vision (Vol. 2352, pp. 19–49). Copenhagen. · Zbl 1039.68686
[41] Mercier, G., & Lennon, M. (2002). On the characterization of hyperspectral texture. In Proceedings of the IEEE international geoscience and remote sensing symposium: Vol. 5. (pp. 2584–2586). Toronto.
[42] Micak, M., Kozintsev, I., Ramchandran, K., & Moulin, P. (1999). Low-complexity image denoising based on statistical modeling of wavelet coefficients. IEEE Signal Processing Letters, 6(12), 300–303. · doi:10.1109/97.803428
[43] MIT vision and modeling group (2010). Vision texture. Online at http://vismod.media.mit.edu/vismod/imagery/VisionTexture/ .
[44] Mitchell, A. (1989). The information matrix, skewness tensor and {\(\alpha\)}-connections for the general multivariate elliptic distribution. Annals of the Institute of Statistical Mathematics, 41(2), 289–304. · Zbl 0691.62049 · doi:10.1007/BF00049397
[45] Muller, M. (1959). A note on a method for generating points uniformly on N-dimensional spheres. Communications of the ACM, 2(4), 19–20. · Zbl 0086.11605 · doi:10.1145/377939.377946
[46] Murray, M., & Rice, J. (1993). Differential geometry and statistics. Monographs on statistics and applied probability: Vol. 48. New York: Chapman & Hall. · Zbl 0804.53001
[47] O’Neill, B. (1982). Elementary differential geometry (2nd ed.). New York: Academic Press.
[48] Pižurica, A., & Philips, W. (2006). Estimating the probability of the presence of a signal of interest in multiresolution single- and multiband image denoising. IEEE Transactions on Image Processing, 15(3), 654–665. · Zbl 05453207 · doi:10.1109/TIP.2005.863698
[49] Rao, C. (1945). Information and accuracy attainable in the estimation of statistical parameters. Bulletin of the Calcutta Mathematical Society, 37, 81–89. · Zbl 0063.06420
[50] Scheunders, P., & De Backer, S. (2007). Wavelet denoising of multicomponent images, using Gaussian Scale Mixture models and a noise-free image as priors. IEEE Transactions on Image Processing, 16(7), 1865–1872. · Zbl 05453650 · doi:10.1109/TIP.2007.899598
[51] Skovgaard, L. (1981). A Riemannian geometry of the multivariate normal model (Tech. Rep. 81/3). Statistical Research Unit, Danish Medical Research Council, Danish Social Science Research Council.
[52] Skovgaard, L. (1984). A Riemannian geometry of the multivariate normal model. Scandinavian Journal of Statistics, 11(4), 211–223. · Zbl 0579.62033
[53] Stephens, M. (1964). The testing of unit vectors for randomness. Journal of the American Statistical Association, 59(305), 160–167. · doi:10.1080/01621459.1964.10480709
[54] Synyavskyy, A., Voloshynovskiy, S., & Prudyus, I. (2001). Wavelet-based map image denoising using provably better class of stochastic I.I.D. image models. Facta Universitatis. Series: Electronics and Energetics, 14(3), 375–385.
[55] The Mathworks (2008). Natick, MA. www.mathworks.com .
[56] Tzagkarakis, G., Beferull-Lozano, B., & Tsakalides, P. (2006). Rotation-invariant texture retrieval with gaussianized steerable pyramids. IEEE Transactions on Image Processing, 15(9), 2702–2718. · Zbl 05453577 · doi:10.1109/TIP.2006.877356
[57] Van de Wouwer, G., Scheunders, P., & Van Dyck, D. (1999). Statistical texture characterization from discrete wavelet representations. IEEE Transactions on Image Processing, 8(4), 592–598. · doi:10.1109/83.753747
[58] Varanasi, M. K., & Aazhang, B. (1989). Parametric generalized Gaussian density estimation. The Journal of the Acoustical Society of America, 86(4), 1404–1415. · doi:10.1121/1.398700
[59] Varma, M., & Zisserman, A. (2003). Texture classification: are filter banks necessary. In Proceedings of the IEEE conference on computer vision and pattern recognition (Vol. 2, pp. 691–698).
[60] Varma, M., & Zisserman, A. (2009). A statistical approach to material classification using image patch exemplars. IEEE Transactions on Pattern Analysis and Machine Intelligence, 31(11), 2032–2047. · doi:10.1109/TPAMI.2008.182
[61] Verdoolaege, G., & Scheunders, P. (2011, submitted). On the geometry of multivariate generalized Gaussian models. Journal of Mathematical Imaging and Vision. · Zbl 1235.68316
[62] Verdoolaege, G., De Backer, S., & Scheunders, P. (2008). Multiscale colour texture retrieval using the geodesic distance between multivariate generalized Gaussian models. In Proceedings of the IEEE international conference on image processing (pp. 169–172). San Diego, CA.
[63] Verdoolaege, G., Rosseel, Y., Lambrechts, M., & Scheunders, P. (2009). Wavelet-based colour texture retrieval using the Kullback-Leibler divergence between bivariate generalized Gaussian models. In Proceedings of the IEEE international conference on image processing (pp. 265–268). Caïro.
[64] Watson, G., & Williams, E. (1956). On the construction of significance tests on the circle and the sphere. Biometrika, 43(3–4), 344–352. · Zbl 0073.14501
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.