Algorithms for Image Processing and Computer Vision

A cookbook of algorithms for common image processing applications Thanks to advances in computer hardware and software, algorithms have been developed that support sophisticated image processing without requiring an extensive background in mathematics. This bestselling book has been fully updated with the newest of these, including 2D vision methods in content-based searches and the use of graphics cards as image processing computational aids. It’s an ideal reference for software engineers and developers, advanced programmers, graphics programmers, scientists, and other specialists who require highly specialized image processing. Algorithms now exist for a wide variety of sophisticated image processing applications required by software engineers and developers, advanced programmers, graphics programmers, scientists, and related specialists This bestselling book has been completely updated to include the latest algorithms, including 2D vision methods in content-based searches, details on modern classifier methods, and graphics cards used as image processing computational aids Saves hours of mathematical calculating by using distributed processing and GPU programming, and gives non-mathematicians the shortcuts needed to program relatively sophisticated applications. Algorithms for Image Processing and Computer Vision, 2nd Edition provides the tools to speed development of image processing applications.

Table of Contents:
Preface xxi Chapter 1 Practical Aspects of a Vision System — Image Display, Input/Output, and Library Calls 1 OpenCV 2 The Basic OpenCV Code 2 The IplImage Data Structure 3 Reading and Writing Images 6 Image Display 7 An Example 7 Image Capture 10 Interfacing with the AIPCV Library 14 Website Files 18 References 18 Chapter 2 Edge-Detection Techniques 21 The Purpose of Edge Detection 21 Traditional Approaches and Theory 23 Models of Edges 24 Noise 26 Derivative Operators 30 Template-Based Edge Detection 36 Edge Models: The Marr-Hildreth Edge Detector 39 The Canny Edge Detector 42 The Shen-Castan (ISEF) Edge Detector 48 A Comparison of Two Optimal Edge Detectors 51 Color Edges 53 Source Code for the Marr-Hildreth Edge Detector 58 Source Code for the Canny Edge Detector 62 Source Code for the Shen-Castan Edge Detector 70 Website Files 80 References 82 Chapter 3 Digital Morphology 85 Morphology Defined 85 Connectedness 86 Elements of Digital Morphology — Binary Operations 87 Binary Dilation 88 Implementing Binary Dilation 92 Binary Erosion 94 Implementation of Binary Erosion 100 Opening and Closing 101 MAX — A High-Level Programming Language for Morphology 107 The ‘‘Hit-and-Miss’’ Transform 113 Identifying Region Boundaries 116 Conditional Dilation 116 Counting Regions 119 Grey-Level Morphology 121 Opening and Closing 123 Smoothing 126 Gradient 128 Segmentation of Textures 129 Size Distribution of Objects 130 Color Morphology 131 Website Files 132 References 135 Chapter 4 Grey-Level Segmentation 137 Basics of Grey-Level Segmentation 137 Using Edge Pixels 139 Iterative Selection 140 The Method of Grey-Level Histograms 141 Using Entropy 142 Fuzzy Sets 146 Minimum Error Thresholding 148 Sample Results From Single Threshold Selection 149 The Use of Regional Thresholds 151 Chow and Kaneko 152 Modeling Illumination Using Edges 156 Implementation and Results 159 Comparisons 160 Relaxation Methods 161 Moving Averages 167 Cluster-Based Thresholds 170 Multiple Thresholds 171 Website Files 172 References 173 Chapter 5 Texture and Color 177 Texture and Segmentation 177 A Simple Analysis of Texture in Grey-Level Images 179 Grey-Level Co-Occurrence 182 Maximum Probability 185 Moments 185 Contrast 185 Homogeneity 185 Entropy 186 Results from the GLCM Descriptors 186 Speeding Up the Texture Operators 186 Edges and Texture 188 Energy and Texture 191 Surfaces and Texture 193 Vector Dispersion 193 Surface Curvature 195 Fractal Dimension 198 Color Segmentation 201 Color Textures 205 Website Files 205 References 206 Chapter 6 Thinning 209 What Is a Skeleton? 209 The Medial Axis Transform 210 Iterative Morphological Methods 212 The Use of Contours 221 Choi/Lam/Siu Algorithm 224 Treating the Object as a Polygon 226 Triangulation Methods 227 Force-Based Thinning 228 Definitions 229 Use of a Force Field 230 Subpixel Skeletons 234 Source Code for Zhang-Suen/Stentiford/Holt Combined Algorithm 235 Website Files 246 References 247 Chapter 7 Image Restoration 251 Image Degradations — The Real World 251 The Frequency Domain 253 The Fourier Transform 254 The Fast Fourier Transform 256 The Inverse Fourier Transform 260 Two-Dimensional Fourier Transforms 260 Fourier Transforms in OpenCV 262 Creating Artificial Blur 264 The Inverse Filter 270 The Wiener Filter 271 Structured Noise 273 Motion Blur — A Special Case 276 The Homomorphic Filter — Illumination 277 Frequency Filters in General 278 Isolating Illumination Effects 280 Website Files 281 References 283 Chapter 8 Classification 285 Objects, Patterns, and Statistics 285 Features and Regions 288 Training and Testing 292 Variation: In-Class and Out-Class 295 Minimum Distance Classifiers 299 Distance Metrics 300 Distances Between Features 302 Cross Validation 304 Support Vector Machines 306 Multiple Classifiers — Ensembles 309 Merging Multiple Methods 309 Merging Type 1 Responses 310 Evaluation 311 Converting Between Response Types 312 Merging Type 2 Responses 313 Merging Type 3 Responses 315 Bagging and Boosting 315 Bagging 315 Boosting 316 Website Files 317 References 318 Chapter 9 Symbol Recognition 321 The Problem 321 OCR on Simple Perfect Images 322 OCR on Scanned Images — Segmentation 326 Noise 327 Isolating Individual Glyphs 329 Matching Templates 333 Statistical Recognition 337 OCR on Fax Images — Printed Characters 339 Orientation — Skew Detection 340 The Use of Edges 345 Handprinted Characters 348 Properties of the Character Outline 349 Convex Deficiencies 353 Vector Templates 357 Neural Nets 363 A Simple Neural Net 364 A Backpropagation Net for Digit Recognition 368 The Use of Multiple Classifiers 372 Merging Multiple Methods 372 Results From the Multiple Classifier 375 Printed Music Recognition — A Study 375 Staff Lines 376 Segmentation 378 Music Symbol Recognition 381 Source Code for Neural Net Recognition System 383 Website Files 390 References 392 Chapter 10 Content-Based Search — Finding Images by Example 395 Searching Images 395 Maintaining Collections of Images 396 Features for Query by Example 399 Color Image Features 399 Mean Color 400 Color Quad Tree 400 Hue and Intensity Histograms 401 Comparing Histograms 402 Requantization 403 Results from Simple Color Features 404 Other Color-Based Methods 407 Grey-Level Image Features 408 Grey Histograms 409 Grey Sigma — Moments 409 Edge Density — Boundaries Between Objects 409 Edge Direction 410 Boolean Edge Density 410 Spatial Considerations 411 Overall Regions 411 Rectangular Regions 412 Angular Regions 412 Circular Regions 414 Hybrid Regions 414 Test of Spatial Sampling 414 Additional Considerations 417 Texture 418 Objects, Contours, Boundaries 418 Data Sets 418 Website Files 419 References 420 Systems 424 Chapter 11 High-Performance Computing for Vision and Image Processing 425 Paradigms for Multiple-Processor Computation 426 Shared Memory 426 Message Passing 427 Execution Timing 427 Using clock() 428 Using QueryPerformanceCounter 430 The Message-Passing Interface System 432 Installing MPI 432 Using MPI 433 Inter-Process Communication 434 Running MPI Programs 436 Real Image Computations 437 Using a Computer Network — Cluster Computing 440 A Shared Memory System — Using the PC Graphics Processor 444 GLSL 444 OpenGL Fundamentals 445 Practical Textures in OpenGL 448 Shader Programming Basics 451 Vertex and Fragment Shaders 452 Required GLSL Initializations 453 Reading and Converting the Image 454 Passing Parameters to Shader Programs 456 Putting It All Together 457 Speedup Using the GPU 459 Developing and Testing Shader Code 459 Finding the Needed Software 460 Website Files 461 References 461 Index 465

About the Author :
J. R. Parker is a full professor working in the Art department at the University of Calgary. His major research projects include live performance in online virtual spaces, the design and construction of kinetic games, and the portrayal of Canadian history and culture in digital and online form.