Abstract
The problem of segmenting an image into several modalities representing different textures can be modelled using Gaussian mixtures. Moreover, texture image patches when translated, rotated or scaled lie in low dimensional subspaces of the high-dimensional space spanned by the grey values. These two aspects make the mixture of local subspace models worth consideration for segmenting this type of images. In recent years a number of mixtures of local PCA models have been proposed. Most of these models require the user to set the number of subspaces and subspace dimensionalities. To make the model autonomous, we propose a greedy EM algorithm to find a suboptimal number of subspaces, besides using a global retained variance ratio to estimate for each subspace the dimensionality that retains the given variability ratio. We provide experimental results for testing the proposed method on texture segmentation.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Cadez, I., Smyth, P.: On model selection and concavity for finite mixture models. In: Symp. On Information theory (ISIT)(2000)
Dasgupta, S.: Learning mixtures of Gaussians. In: Proc. IEEE Symposium on Foundation of Computer Science (1999)
de Ridder, D., Kittler, J., Duin, R.: Probabilistic PCA and ICA subspace mixture models for image segmentation. In: Proc. 11th British Machine Vision Conference (BMVC 2000), Bristol, UK, pp. 112–121 (2000)
de Ridder, D.: Adaptive methods of image processing. Doc. dissertation, Faculty of Applied Science, Delft University of Technology (2001)
Hinton, G., Dayan, P., Revow, M.: Modeling the manifolds of images of handwritten digits. IEEE Trans. on Neural Networks. 10(3), 65–74 (1997)
Kambhatla, N.: Local models & Gaussian mixture models for statistical data processing. Doc. dissertation, Oregon Graduate Inst. of Science & Tech. (1995)
Li, J., Barron, A.: Mixture density estimation. In: Advances in Neural Inf. Proc. Systems, vol. 12, MIT Press, Cambridge (2000)
Meinicke, P., Ritter, H.: Resolution-based Complexity control for Gaussian mixture models. Neural Comp. 13(2), 453–475 (2001)
Moghaddam, B., Pentland, A.: Probabilistic visual learning for object representation. IEEE Trans. PAMI 19(7), 696–710 (1997)
Musa, M., Duin, R., de Ridder, D.: An enhanced EM algorithm for mixture of probabilistic principal component analysis. In: ICANN 2001 Workshop on Kernel & subspace Methods for computer Vision (2001)
Musa, M., Duin, R., de Ridder, D.: Almost automomous training of mixtures of principal component analyzers. Submittend to Pattern Rec. Let. (2003)
Tipping, M., Bishop, C.: Mixtures of principal component analyzers. Neural Comp. 11(2), 443–482 (1999)
Verbeek, J., Vlassis, N., Krose, B.: Efficient greedy learning of Gaussian mixture models. Neural Comp. 15(2), 469–485 (2003)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Musa, M.E.M., Duin, R.P.W., de Ridder, D., Atalay, V. (2003). Texture Segmentation Using the Mixtures of Principal Component Analyzers. In: Yazıcı, A., Şener, C. (eds) Computer and Information Sciences - ISCIS 2003. ISCIS 2003. Lecture Notes in Computer Science, vol 2869. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-39737-3_63
Download citation
DOI: https://doi.org/10.1007/978-3-540-39737-3_63
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-20409-1
Online ISBN: 978-3-540-39737-3
eBook Packages: Springer Book Archive