Cluster-based probability model and its application to image and texture processing

@article{Popat1997ClusterbasedPM,
  title={Cluster-based probability model and its application to image and texture processing},
  author={Ashok Popat and Rosalind W. Picard},
  journal={IEEE transactions on image processing : a publication of the IEEE Signal Processing Society},
  year={1997},
  volume={6 2},
  pages={
          268-84
        }
}
  • Ashok Popat, Rosalind W. Picard
  • Published 1 February 1997
  • Computer Science
  • IEEE transactions on image processing : a publication of the IEEE Signal Processing Society
We develop, analyze, and apply a specific form of mixture modeling for density estimation within the context of image and texture processing. The technique captures much of the higher order, nonlinear statistical relationships present among vector elements by combining aspects of kernel estimation and cluster analysis. Experimental results are presented in the following applications: image restoration, image and texture compression, and texture classification. 
View on PubMed
doi.org
152 Citations
Highly Influential Citations
8
Background Citations
33
Methods Citations
36
Results Citations
1

Figures from this paper

152 Citations

Color image edge detection based on nonparametric density estimation
TLDR
A novel computationally efficient adaptive algorithm to accomplish edge detection in multidimensional and color images is presented, based on local, non-parametric kernel density estimation.
A cluster-based statistical model for object detection
TLDR
This paper presents an approach to object detection which is based on recent work in statistical models for texture synthesis and recognition, and presents promising results in applying the technique to face detection and car detection.
Mixture models and the segmentation of multimodal textures
  • R. Manduchi
  • Computer Science
    Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662)
  • 2000
TLDR
A divide-and-conquer method that groups together Gaussian clusters (estimated via Expectation Maximization) into homogeneous texture classes that allows to successfully segment even rather complex textures, as demonstrated by experimental tests on natural images.
Non-Gaussian mixture image models prediction
  • N. Bouguila
  • Computer Science
    2008 15th IEEE International Conference on Image Processing
  • 2008
TLDR
The generalized Dirichlet mixture-based predictor is nonlinear and takes into account the fact that images clutter and texture are generally non-Gaussian.
A Texture-Based Statistical Model for Face Detection
TLDR
A general statistical model which relies on texture for learning an object class from a set of example images that can handle larger variations in the appearance of faces than previous approaches is proposed.
Texture characterization via joint statistics of wavelet coefficient magnitudes
TLDR
An efficient algorithm for sampling from an implicit probability density conforming to these statistics is developed, and its effectiveness in synthesizing artificial and natural texture images is demonstrated.
A Parametric Texture Model Based on Joint Statistics of Complex Wavelet Coefficients
TLDR
A universal statistical model for texture images in the context of an overcomplete complex wavelet transform is presented, demonstrating the necessity of subgroups of the parameter set by showing examples of texture synthesis that fail when those parameters are removed from the set.
Higher-order image statistics for unsupervised, information-theoretic, adaptive, image filtering
  • Suyash P. Awate, R. Whitaker
  • Computer Science
    2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
  • 2005
TLDR
A novel unsupervised, information-theoretic, adaptive filter (UINTA) that improves the predictability of pixel intensities from their neighborhoods by decreasing the joint entropy between them and can thereby restore a wide spectrum of images and applications.
Statistical models for images: compression, restoration and synthesis
  • Eero P. Simoncelli
  • Computer Science
    Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)
  • 1997
TLDR
A parametric statistical model for visual images in the wavelet transform domain is presented and the joint densities of coefficient magnitudes at adjacent spatial locations, adjacent orientations, and adjacent spatial scales are characterized.
Variable Bandwidth Image Models for Texture-Preserving Enhancement of Natural Images
  • N. Azzabou
  • Environmental Science, Computer Science
  • 2008
TLDR
This thesis introduces novel non-parametric image models towards more appropriate and adaptive image description using variable bandwidth approximations driven from a soft classification in the image and introduces alternative means to model observations dependencies from geometric point of view.
...
...

References

SHOWING 1-10 OF 49 REFERENCES
Novel cluster-based probability model for texture synthesis, classification, and compression
TLDR
A new probabilistic modeling technique for high-dimensional vector sources is presented, and its application to the problems of texture synthesis, classification, and compression is considered.
Simultaneous learning of decision rules and important attributes for classification problems in image analysis
Density Estimation for Statistics and Data Analysis.
TLDR
The Kernel Method for Multivariate Data: Three Important Methods and Density Estimation in Action.
Networks for approximation and learning
The problem of the approximation of nonlinear mapping, (especially continuous mappings) is considered. Regularization theory and a theoretical framework for approximation (based on regularization…
Bayes risk weighted vector quantization with posterior estimation for image compression and classification
TLDR
This work investigates several VQ-based algorithms that seek to minimize both the distortion of compressed images and errors in classifying their pixel blocks and introduces a tree-structured posterior estimator to produce the class posterior probabilities required for the Bayes risk computation in this design.
Optimal nonlinear interpolative vector quantization
  • A. Gersho
  • Computer Science
    IEEE Trans. Commun.
  • 1990
TLDR
The range of applicability of nonlinear interpolative vector quantization is illustrated with examples in which optimal nonlinear estimation from quantized data is needed for efficient signal compression.
Multivariate Density Estimation
Exploring and identifying structure is even more important for multivariate data than univariate data, given the difficulties in graphically presenting multivariate data and the comparative lack of…
On the JPEG model for lossless image compression
TLDR
In this study, the Bostelmann (1974) technique is studied for use at all resolutions, whereas in the arithmetic coded JPEG lossless, the technique is applied only at the 16-bit per pixel resolution.
An Algorithm for Vector Quantizer Design
An efficient and intuitive algorithm is presented for the design of vector quantizers based either on a known probabilistic model or on a long training sequence of data. The basic properties of the…
Fast Learning in Networks of Locally-Tuned Processing Units
We propose a network architecture which uses a single internal layer of locally-tuned processing units to learn both classification tasks and real-valued function approximations (Moody and Darken…
...
...