图像处理、分析与机器视觉 英文版PDF|Epub|txt|kindle电子书版本网盘下载

图像处理、分析与机器视觉 英文版PDF格式电子书版下载

注意：本站所有压缩包均有解压码： 点击下载压缩包解压工具

1 Introduction1

1.1 Summary8

1.2 Exercises8

1.3 References9

2 The digitized image and its properties10

2.1 Basic concepts10

2.1.1 Image functions10

2.1.3 The Fourier transform13

2.1.2 The Dirac distribution and convolution13

2.1.4 Images as a stochastic process15

2.1.5 Images as linear systems17

2.2 Image digitization18

2.2.1 Sampling18

2.2.2 Quantization22

2.2.3 Color images23

2.3 Digital image properties27

2.3.1 Metric and topological properties of digital images27

2.3.2 Histograms32

2.3.3 Visual perception of the image33

2.3.4 Image quality35

2.3.5 Noise in images35

2.4 Summary37

2.5 Exercises38

2.6 References40

3 Data structures for image analysis42

3.1 Levels of image data representation42

3.2 Traditional image data structures43

3.2.1 Matrices43

3.2.2 Chains45

3.2.3 Topological data structures47

3.2.4 Relational structures48

3.3 Hierarchical data structures49

3.3.1 Pyamids49

3.3.2 Quadtrees51

3.3.3 Other pyramidical structures52

3.4 Summary53

3.5 Exercises54

3.6 References55

4 Image pre-processing57

4.1 Pixel brightness transformations58

4.1.1 Position-dependent brightness correction58

4.1.2 Gray-scale transformation59

4.2 Geometric transformations62

4.2.1 Pixel co-ordinate transformations63

4.2.2 Brightness interpolation65

4.3 Local pre-processing68

4.3.1 Image smoothing69

4.3.2 Edge detectors77

4.3.3 Zero-crossings of the second derivative83

4.3.4 Scale in image processing88

4.3.5 Canny edge detection90

4.3.6 Parametric edge models93

4.3.7 Edges in multi-spectral images94

4.3.8 Other local pre-processing operators94

4.3.9 Adaptive neighborthood pre-processing98

4.4 Image restoration102

4.4.1 Degradations that are easy to restore105

4.4.3 Wiener filtration106

4.4.2 Inverse filtration106

4.5 Summary108

4.6 Exercises111

4.7 References118

5 Segmentation123

5.1 Thresholding124

5.1.1 Threshold detection methods127

5.1.2 Optimal thresholding128

5.1.3 Multi-spectral thresholding131

5.1.4 Thresholding in hierarchical data structures133

5.2 Edge-based segmentation134

5.2.1 Edge image thresholding135

5.2.2 Edge relaxation137

5.2.3 Border tracing142

5.2.4 Border detection as graph searching148

5.2.5 Border detection as dynamic programming158

5.2.6 Hough transforms163

5.2.7 Border detection using border location information173

5.2.8 Region construction form borders174

5.3 Region-based segmentation176

5.3.1 Region merging177

5.3.2 Region splitting181

5.3.3 Splitting and merging181

5.3.4 Watershed segmentation186

5.3.5 Region growing post-processing188

5.4 Matching190

5.4.1 Matching criteria191

5.4.2 Control strategies of matching193

5.5.1 Simultaneous detection of border pairs194

5.5 Advanced optimal border and surface detection approaches194

5.5.2 Surface detection199

5.6 Summary205

5.7 Exercises210

5.8 References216

6 Shape representation and description228

6.1 Region identification232

6.2 Contour-based shape representation and description235

6.2.1 Chain codes236

6.2.2 Simple geometric border representation237

6.2.3 Fourier transforms of boundaries240

6.2.4 Boundary description using segment sequences242

6.2.5 B-spline representation245

6.2.6 Other contour-based shape description approaches248

6.2.7 Spape invariants249

6.3 Region-based shape representation and description254

6.3.1 Simple scalar region descriptors254

6.3.2 Moments259

6.3.3 Convex hull262

6.3.4 Graph representation based on region skeleton267

6.3.5 Region decomposition271

6.3.6 Region neighborhood graphs272

6.4 Shape classes273

6.5 Summary274

6.6 Exercises276

6.7 References279

7 Object recognition290

7.1 Knowledge representation291

7.2 Statistical pattern recognition297

7.2.1 Classification principles298

7.2.2 Classifier setting300

7.2.3 Classifier learning303

7.2.4 Cluster analysis307

7.3 Neural nets308

7.3.1 Feed-forward networks310

7.3.2 Unsupervised learning312

7.3.3 Hopfield neural nets313

7.4 Syntactic pattern recognition315

7.4.1 Grammars and languages317

7.4.2 Syntactic analysis, syntactic classifier319

7.4.3 Syntactic classifier learning, grammar inference321

7.5 Recognition as graph matching323

7.5.1 Isomorphism of graphs and sub-graphs324

7.5.2 Similarity of graphs328

7.6 Optimization techniques in recognition328

7.6.1 Genetic algorithms330

7.6.2 Simulated annealing333

7.7.1 Fuzzy sets and fuzzy membership functions336

7.7 Fuzzy systems336

7.7.2 Fuzzy set operators338

7.7.3 Fuzzy reasoning339

7.7.4 Fuzzy system design and training343

7.8 Summary344

7.9 Exercises347

7.10 References354

8 Image understanding362

8.1.2 Hierarchical control364

8.1.1 Parallel and serial processing control364

8.1 Image understanding control strategies364

8.1.3 Bottom-up control strategies365

8.1.4 Model-based control strategies366

8.1.5 Combined control strategies367

8.1.6 Non-hierarchical control371

8.2 Active contour models-snakes374

8.3 Point distribution models380

8.4 Pattern recognition methods in image understanding390

8.4.1 Contextual image classification392

8.5 Scene labeliing and constraint propagation397

8.5.1 Discrete relaxation398

8.5.2 Probabilistic relaxation400

8.5.3 Searching interpretation trees404

8.6 Semantic image segmentation and understanding404

8.6.1 Semantic region growing406

8.6.2 Genetic image interpretation408

8.7 Hidden Markov models417

8.8 Summary423

8.9 Exercises426

8.10 References428

9 3D vision, geometry, and radiometry441

9.1 3D vision tasks442

9.1.1 Marr s theory444

9.1.2 Other vision paradigms: Active and purposive vision446

9.2 Geometry for 3D vision448

9.2.1 Basic of projective geometry448

9.2.2 The single perspective camera449

9.2.3 An overview of single camera calibration453

9.2.4 Calibration of one camera from a known scene455

9.2.5 Two cameras, stereopsis457

9.2.6 The geometry of two cameras；the fundamental matrix460

9.2.7 Relative motion of the camera；the essential matrix462

9.2.8 Fundamental matrix estimation from image point correspondences464

9.2.9 Applications of epipolar geometry in vision466

9.2.10 Three and more cameras471

9.2.11 Stereo correspondence algorithms476

9.2.12 Active acquisition of range images483

9.3 Radiometry and 3D vision486

9.3.1 Radiometric considerations in determining gary-level486

9.3.2 Surface reflectance490

9.3.3 Shape from shading494

9.3.4 Photometric stereo498

9.4 Summary499

9.5 Exercises501

9.6 References502

10 Use of 3D vision508

10.1 Shape from X508

10.1.1 Shape from motion508

10.1.2 Shape from texture515

10.1.3 Other shape from X techniques517

10.2 Full 3D objects519

10.2.1 3D objects, models, and related issues519

10.2.2 Line labeling521

10.2.3 Volumetirc representation, direct mesurements523

10.2.4 Volumetric modeling strategies525

10.2.5 Surface modeling strategies527

10.2.6 Registering surface patches and their fusion to get a full 3D model529

10.3.1 General considerations535

10.3 3D model-based vision535

10.3.2 Goad s algorithm537

10.3.3 Model-based recognition of curved objects from intensity images541

10.3.4 Model-based recognition based on range images543

10.4 2D view-based representations of a 3D scene544

10.4.1 Viewing space544

10.4.2 Multi-view representations and aspect graphs544

10.4.3 Geons as a 2D view-based structural representation545

10.4.4 Visualizing 3D real-world scenes using stored collections of 2D view546

10.5 Summary551

10.6 Exercies552

10.7 References553

11 Mathematical morphology559

11.1 Basic morphological concepts559

11.2 Four morphological principles561

11.3 Binary dilation and erosion563

11.3.1 Dilation563

11.3.2 Erosion565

11.3.3 Hit-or-miss transformation568

11.3.4 Opening and closing568

11.4 Gray-scle dilation and erosion569

11.4.1 Top surface, umbra, and gray-scale dilation and erosion570

11.4.2 Umbra homeomorphism theorem, properties of erosion and dilation,opening and closing573

11.4.3 Top hat transformation574

11.5 Skeltons and object marking576

11.5.1 Homotopic transformations576

11.5.2 Skeletion, maximal ball576

11.5.3 Thinning, thickening,and homotopic skeleton578

11.5.4 Quench function, ultimate erosion581

11.5.5 Ultimate erosion and distance functions584

11.5.6 Geodesic trandformations585

11.5.7 Morphological reconstruction586

11.6 Granulometry589

11.7 Morphological segmentation and watersheds590

11.7.1 Particles segmenttation, marking, and watersheds590

11.7.2 Binary morphological segmentation592

11.7.3 Gary-scale segmentation, watersheds594

11.8 Summary595

11.9 Exercises597

11.10 References598

12 Linear discrete image tranforms600

12.1 Basic theory600

12.2 Fourier transform602

12.3 Hadamard transform604

12.4 Discrete cosine transform605

12.5 Wavelets606

12.6 Other orthogonal image transforms608

12.7 Applications of discrete image transforms609

12.8 Summary613

12.9 Exercises617

12.10 References619

13.Image data compression621

13.1 Image data properties622

13.2 Discrete image tranforms in image data compression623

13.3 Predictive compression methods624

13.4 Vector quantization629

13.5 Hierarchical and progressive compression methods630

13.6 Comparison of compression methods631

13.7 Other techniques632

13.8 Coding633

13.9 JPEG and MPEG image compression634

13.9.1 JPEG-still image compression634

13.9.2 MPEG-full-motion video compression636

13.10 Summary637

13.11 Exercises640

13.12 References641

14 Texture646

14.1 Statistical texture description649

14.1.1 Methods based on spatial frequencies649

14.1.2 Co-occurrence matices651

14.1.3 Edge frequency653

14.1.4 Primitive length (run length)655

14.1.5 Laws texture energy measures656

14.1.6 Fractal texture description657

14.1.7 Other statistical methods of texture description659

14.2 Syntactic texture description methods660

14.2.1 Shape chain grammars661

14.2.2 Graph grammars663

14.2.3 Primitive grouping in hierarchical textures664

14.3 Hybrid texture description methods666

14.4 Texture recognition method applications667

14.5 Summary668

14.6 Exercises670

14.7 References672

15 Motion analysis679

15.1 Differential mition analysis methods682

15.2 Optical flow685

15.2.1 Optical flow computation686

15.2.2 Global and local optical fow estimation689

15.2.3 Optical flow computation approaches690

15.2.4 Optical flow in motion analysis693

15.3 Analysis based on correspondence of interest points696

15.3.1 Detection of interest points696

15.3.2 Correspondence of interest points697

15.3.3 Object tracking700

15.4 Kalman filters708

15.4.1 Example709

15.5 Summary710

15.6 Exercises712

15.7 References714

16 Case studies722

16.1 An optical music recognition system722

16.2 Automated image analysis in cardiology727

16.2.1 Robust analysis of coronry angiograms730

16.2.2 Knowledge-based analysis of intra-vascular ultrasound733

16.3 Automated indentification of airway trees738

16.4 Passive surveillance744

16.5 References750

Index755