Pattern Recognition and Image Preprocessing

by Bow; Sing T.

• ISBN13: 9780824706593
• ISBN10: 0824706595
• Edition: 2nd
• Format: Hardcover
• Copyright: 2002-01-11
• Publisher: CRC Press

Summary

Describing non-parametric and parametric theoretic classification and the training of discriminant functions, this second edition includes new and expanded sections on neural networks, Fisher's discriminant, wavelet transform, and the method of principal components. It contains discussions on dimensionality reduction and feature selection, novel computer system architectures, proven algorithms for solutions to common roadblocks in data processing, computing models including the Hamming net, the Kohonen self-organizing map, and the Hopfield net, detailed appendices with data sets illustrating key concepts in the text, and more.

Author Biography

Sing-Tze Bow is Professor, Department of Electrical Engineering, Northern Illinois University, De Kalb.

Table of Contents

Series Introduction	p. iii
Preface	p. v
Pattern Recognition	p. 1
Introduction	p. 3
Patterns and Pattern Recognition	p. 3
Significance and Potential Function of the Pattern Recognition System	p. 5
Configuration of the Pattern Recognition System	p. 8
Representation of Patterns and Approaches to Their Machine Recognition	p. 16
Paradigm Applications	p. 23
Supervised and Unsupervised Learning in Pattern Recognition	p. 29
Nonparametric Decision Theoretic Classification	p. 33
Decision Surfaces and Discriminant Functions	p. 34
Linear Discriminant Functions	p. 38
Piecewise Linear Discriminant Functions	p. 42
Nonlinear Discriminant Functions	p. 49
[phi] Machines	p. 52
Potential Functions as Discriminant Functions	p. 57
Problems	p. 59
Nonparametric (Distribution-Free) Training of Discriminant Functions	p. 62
Weight Space	p. 62
Error Correction Training Procedures	p. 66
Gradient Techniques	p. 72
Training Procedures for the Committee Machine	p. 74
Practical Considerations Concerning Error Correction Training Methods	p. 76
Minimum-Squared-Error Procedures	p. 76
Problems	p. 79
Statistical Discriminant Functions	p. 82
Introduction	p. 82
Problem Formulation by Means of Statistical Design Theory	p. 83
Optimal Discriminant Functions for Normally Distributed Patterns	p. 93
Training for Statistical Discriminant Functions	p. 101
Application to a Large Data-Set Problem: A Practical Example	p. 102
Problems	p. 106
Clustering Analysis and Unsupervised Learning	p. 112
Introduction	p. 112
Clustering with an Unknown Number of Classes	p. 117
Clustering with a Known Number of Classes	p. 129
Evaluation of Clustering Results by Various Algorithms	p. 145
Graph Theoretical Methods	p. 146
Mixture Statistics and Unsupervised Learning	p. 161
Concluding Remarks	p. 164
Problems	p. 164
Dimensionality Reduction and Feature Selection	p. 168
Optimal Number of Features in Classification of Multivariate Gaussian Data	p. 168
Feature Ordering by Means of Clustering Transformation	p. 170
Canonical Analysis and Its Applications to Remote Sensing Problems	p. 172
Optimum Classification with Fisher's Discriminant	p. 182
Nonparametric Feature Selection Method Applicable to Mixed Features	p. 188
Problems	p. 190
Neural Networks for Pattern Recognition	p. 197
Multilayer Perceptron	p. 201
Some Preliminaries	p. 201
Pattern Mappings in a Multilayer Perceptron	p. 205
A Primitive Example	p. 219
Problems	p. 223
Radial Basis Function Networks	p. 225
Radial Basis Function Networks	p. 225
RBF Network Training	p. 231
Formulation of the Radial Basis Functions for Pattern Classification by Means of Statistical Decision Theory	p. 232
Comparison of RBF Networks with Multilayer Perceptrons	p. 234
Problems	p. 235
Hamming Net and Kohonen Self-Organizing Feature Map	p. 236
Hamming Net	p. 236
Kohonen Self-Organizing Feature Map	p. 246
Problems	p. 253
The Hopfield Model	p. 256
The Hopfield Model	p. 256
An Illustrative Example for the Explanation of the Hopfield Network Operation	p. 258
Operation of the Hopfield Network	p. 261
Problems	p. 267
Data Preprocessing for Pictorial Pattern Recognition	p. 269
Preprocessing in the Spatial Domain	p. 271
Deterministic Gray-Level Transformation	p. 271
Gray-Level Histogram Modification	p. 274
Smoothing and Noise Elimination	p. 298
Edge Sharpening	p. 303
Thinning	p. 333
Morphological Processing	p. 336
Boundary Detection and Contour Tracing	p. 343
Texture and Object Extraction from Textural Background	p. 352
Problems	p. 357
Pictorial Data Preprocessing and Shape Analysis	p. 363
Data Structure and Picture Representation by a Quadtree	p. 363
Dot-Pattern Processing with Voronoi Approach	p. 365
Encoding of a Planar Curve by Chain Code	p. 374
Polygonal Approximation of Curves	p. 374
Encoding of a Curve with B-Spline	p. 377
Shape Analysis via Medial Axis Transformation	p. 378
Shape Discrimination Using Fourier Descriptor	p. 380
Shape Description via the Use of Critical Points	p. 384
Shape Description via Concatenated Arcs	p. 385
Identification of Partially Obscured Objects	p. 392
Recognizing Partially Occluded Parts by the Concept of Saliency of a Boundary Segment	p. 394
Problems	p. 399
Transforms and Image Processing in the Transform Domain	p. 401
Formulation of the Image Transform	p. 403
Functional Properties of the Two-Dimensional Fourier Transform	p. 406
Sampling	p. 420
Fast Fourier Transform	p. 437
Other Image Transforms	p. 454
Enhancement by Transform Processing	p. 468
Problems	p. 476
Wavelets and Wavelet Transform	p. 481
Introduction	p. 481
Wavelets and Wavelet Transform	p. 484
Scaling Function and Wavelet	p. 486
Filters and Filter Banks	p. 490
Digital Implementation of DWT	p. 496
Applications	p. 509
Exemplary Applications	p. 511
Document Image Analysis	p. 513
Industrial Inspection	p. 529
Remote Sensing Applications	p. 545
Vision Used for Control	p. 551
Practical Concerns of Image Processing and Pattern Recognition	p. 561
Computer System Architectures for Image Processing and Pattern Recognition	p. 563
What We Expect to Achieve from the Point of View of Computer System Architecture	p. 563
Overview of Specific Logic Processing and Mathematical Computation	p. 564
Interconnection Networks for SIMD Computers	p. 566
Systolic Array Architecture	p. 566
Digitized Images	p. 573
Image Model and Discrete Mathematics	p. 579
Image Model	p. 579
Simplification of the Continuous Image Model	p. 581
Two-Dimensional Delta Function	p. 584
Additive Linear Operators	p. 586
Convolution	p. 587
Differential Operators	p. 590
Preliminaries of Some Methods Used Frequently in Image Preprocessing	p. 591
Problems	p. 595
Digital Image Fundamentals	p. 597
Sampling and Quantization of an Image	p. 597
Imaging Geometry	p. 599
Matrix Manipulation	p. 613
Definition	p. 613
Matrix Multiplication	p. 614
Partitioning of Matrices	p. 615
Computation of the Inverse Matrix	p. 617
Eigenvectors and Eigenvalues of an Operator	p. 621
Notation	p. 625
Bibliography	p. 645
Index	p. 691
Table of Contents provided by Syndetics. All Rights Reserved.