Treiber An Introduction to Object Recognition

Selected Algorithms for a Wide Variety of Applications
1. Auflage 2010
ISBN: 978-1-84996-235-3
Verlag: Springer
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 202 Seiten

Reihe: Advances in Computer Vision and Pattern Recognition

Rapid development of computer hardware has enabled usage of automatic object recognition in an increasing number of applications, ranging from industrial image processing to medical applications, as well as tasks triggered by the widespread use of the internet. Each area of application has its specific requirements, and consequently these cannot all be tackled appropriately by a single, general-purpose algorithm. This easy-to-read text/reference provides a comprehensive introduction to the field of object recognition (OR). The book presents an overview of the diverse applications for OR and highlights important algorithm classes, presenting representative example algorithms for each class. The presentation of each algorithm describes the basic algorithm flow in detail, complete with graphical illustrations. Pseudocode implementations are also included for many of the methods, and definitions are supplied for terms which may be unfamiliar to the novice reader. Supporting a clear and intuitive tutorial style, the usage of mathematics is kept to a minimum. Topics and features: presents example algorithms covering global approaches, transformation-search-based methods, geometrical model driven methods, 3D object recognition schemes, flexible contour fitting algorithms, and descriptor-based methods; explores each method in its entirety, rather than focusing on individual steps in isolation, with a detailed description of the flow of each algorithm, including graphical illustrations; explains the important concepts at length in a simple-to-understand style, with a minimum usage of mathematics; discusses a broad spectrum of applications, including some examples from commercial products; contains appendices discussing topics related to OR and widely used in the algorithms, (but not at the core of the methods described in the chapters). Practitioners of industrial image processing will find this simple introduction and overview to OR a valuable reference, as will graduate students in computer vision courses. Marco Treiber is a software developer at Siemens Electronics Assembly Systems, Munich, Germany, where he is Technical Lead in Image Processing for the Vision System of SiPlace placement machines, used in SMT assembly.

Autoren/Hrsg.

Treiber, Marco Alexander

Inhaltsverzeichnis

1;Preface;6
2;Acknowledgments;9
3;Contents;10
4;Abbreviations;14
5;1 Introduction;15
5.1;1.1 Overview;15
5.2;1.2 Areas of Application;17
5.3;1.3 Requirements and Constraints;18
5.4;1.4 Categorization of Recognition Methods;21
5.5;References;24
6;2 Global Methods;25
6.1;2.1 2D Correlation;25
6.1.1;2.1.1 Basic Approach;25
6.1.1.1;2.1.1.1 Main Idea;25
6.1.1.2;2.1.1.2 Example;27
6.1.1.3;2.1.1.3 Pseudocode;28
6.1.1.4;2.1.1.4 Rating;28
6.1.2;2.1.2 Variants;29
6.1.2.1;2.1.2.1 Variant 1: Preprocessing;29
6.1.2.2;2.1.2.2 Variant 2: Subsampling/Image Pyramids;31
6.1.3;2.1.3 Phase-Only Correlation (POC);32
6.1.3.1;2.1.3.1 Example;33
6.1.3.2;2.1.3.2 Pseudocode;34
6.1.4;2.1.4 Shape-Based Matching;34
6.1.4.1;2.1.4.1 Main Idea;34
6.1.4.2;2.1.4.2 Example;35
6.1.4.3;2.1.4.3 Pseudocode;35
6.1.4.4;2.1.4.4 Rating;36
6.1.5;2.1.5 Comparison;36
6.2;2.2 Global Feature Vectors;38
6.2.1;2.2.1 Main Idea;38
6.2.2;2.2.2 Classification;38
6.2.3;2.2.3 Rating;39
6.2.4;2.2.4 Moments;39
6.2.4.1;2.2.4.1 Main Idea;39
6.2.4.2;2.2.4.2 Example;40
6.2.5;2.2.5 Fourier Descriptors;41
6.2.5.1;2.2.5.1 Main Idea;41
6.2.5.2;2.2.5.2 Example;41
6.2.5.3;2.2.5.3 Modifications;43
6.2.5.4;2.2.5.4 Pseudocode;44
6.3;2.3 Principal Component Analysis (PCA);45
6.3.1;2.3.1 Main Idea;45
6.3.2;2.3.2 Pseudocode;48
6.3.3;2.3.3 Rating;49
6.3.4;2.3.4 Example;49
6.3.5;2.3.5 Modifications;51
6.4;References;52
7;3 Transformation-Search Based Methods;54
7.1;3.1 Overview;54
7.2;3.2 Transformation Classes;55
7.3;3.3 Generalized Hough Transform;57
7.3.1;3.3.1 Main Idea;57
7.3.2;3.3.2 Training Phase;57
7.3.3;3.3.3 Recognition Phase;58
7.3.4;3.3.4 Pseudocode;59
7.3.5;3.3.5 Example;60
7.3.6;3.3.6 Rating;62
7.3.7;3.3.7 Modifications;63
7.4;3.4 The Hausdorff Distance;64
7.4.1;3.4.1 Basic Approach;64
7.4.1.1;3.4.1.1 Main Idea;64
7.4.1.2;3.4.1.2 Recognition Phase;65
7.4.1.3;3.4.1.3 Pseudocode;68
7.4.1.4;3.4.1.4 Example;70
7.4.1.5;3.4.1.5 Rating;71
7.4.2;3.4.2 Variants;72
7.4.2.1;3.4.2.1 Variant 1: Generalized Hausdorff Distance Generalized Hausdorff distance ;72
7.4.2.2;3.4.2.2 Variant 2: 3D Hausdorff Distance;72
7.4.2.3;3.4.2.3 Variant 3: Chamfer Matching;73
7.5;3.5 Speedup by Rectangular Filters and Integral Images;73
7.5.1;3.5.1 Main Idea;73
7.5.2;3.5.2 Filters and Integral Images;74
7.5.3;3.5.3 Classification;76
7.5.4;3.5.4 Pseudocode;78
7.5.5;3.5.5 Example;79
7.5.6;3.5.6 Rating;80
7.6;References;80
8;4 Geometric Correspondence-Based Approaches;82
8.1;4.1 Overview;82
8.2;4.2 Feature Types and Their Detection;83
8.2.1;4.2.1 Geometric Primitives;84
8.2.1.1;4.2.1.1 Polygonal Approximation;84
8.2.1.2;4.2.1.2 Approximation with Line Segments and Circular Arcs;84
8.2.2;4.2.2 Geometric Filters;87
8.3;4.3 Graph-Based Matching;88
8.3.1;4.3.1 Geometrical Graph Match;88
8.3.1.1;4.3.1.1 Main Idea;88
8.3.1.2;4.3.1.2 Recognition Phase;89
8.3.1.3;4.3.1.3 Pseudocode;91
8.3.1.4;4.3.1.4 Example;92
8.3.1.5;4.3.1.5 Rating;92
8.3.2;4.3.2 Interpretation Trees;93
8.3.2.1;4.3.2.1 Main Idea;93
8.3.2.2;4.3.2.2 Recognition Phase;94
8.3.2.3;4.3.2.3 Pseudocode;97
8.3.2.4;4.3.2.4 Example;98
8.3.2.5;4.3.2.5 Rating;99
8.4;4.4 Geometric Hashing;100
8.4.1;4.4.1 Main Idea;100
8.4.2;4.4.2 Speedup by Pre-processing;101
8.4.3;4.4.3 Recognition Phase;102
8.4.4;4.4.4 Pseudocode;103
8.4.5;4.4.5 Rating;104
8.4.6;4.4.6 Modifications;104
8.5;References;105
9;5 Three-Dimensional Object Recognition;107
9.1;5.1 Overview;107
9.2;5.2 The SCERPO System: Perceptual Grouping;109
9.2.1;5.2.1 Main Idea;109
9.2.2;5.2.2 Recognition Phase;110
9.2.3;5.2.3 Example;111
9.2.4;5.2.4 Pseudocode;111
9.2.5;5.2.5 Rating;112
9.3;5.3 Relational Indexing;113
9.3.1;5.3.1 Main Idea;113
9.3.2;5.3.2 Teaching Phase;114
9.3.3;5.3.3 Recognition Phase;116
9.3.4;5.3.4 Pseudocode;117
9.3.5;5.3.5 Example;118
9.3.6;5.3.6 Rating;120
9.4;5.4 LEWIS: 3D Recognition of Planar Objects;120
9.4.1;5.4.1 Main Idea;120
9.4.2;5.4.2 Invariants;121
9.4.3;5.4.3 Teaching Phase;123
9.4.4;5.4.4 Recognition Phase;124
9.4.5;5.4.5 Pseudocode;125
9.4.6;5.4.6 Example;126
9.4.7;5.4.7 Rating;127
9.5;References;128
10;6 Flexible Shape Matching;129
10.1;6.1 Overview;129
10.2;6.2 Active Contour Models/Snakes;130
10.2.1;6.2.1 Standard Snake;130
10.2.1.1;6.2.1.1 Main Idea;130
10.2.1.2;6.2.1.2 Optimization;131
10.2.1.3;6.2.1.3 Example;132
10.2.1.4;6.2.1.4 Rating;133
10.2.2;6.2.2 Gradient Vector Flow Snake;134
10.2.2.1;6.2.2.1 Main Idea;134
10.2.2.2;6.2.2.2 Pseudocode;135
10.2.2.3;6.2.2.3 Example;136
10.2.2.4;6.2.2.4 Rating;137
10.3;6.3 The Contracting Curve Density Algorithm (CCD);138
10.3.1;6.3.1 Main Idea;138
10.3.2;6.3.2 Optimization;140
10.3.3;6.3.3 Example;141
10.3.4;6.3.4 Pseudocode;142
10.3.5;6.3.5 Rating;142
10.4;6.4 Distance Measures for Curves;143
10.4.1;6.4.1 Turning Functions;143
10.4.1.1;6.4.1.1 Main Idea;143
10.4.1.2;6.4.1.2 Example;145
10.4.1.3;6.4.1.3 Pseudocode;146
10.4.1.4;6.4.1.4 Rating;147
10.4.2;6.4.2 Curvature Scale Space (CSS);147
10.4.2.1;6.4.2.1 Main Idea;147
10.4.2.2;6.4.2.2 Pseudocode;150
10.4.2.3;6.4.2.3 Rating;151
10.4.3;6.4.3 Partitioning into Tokens;151
10.4.3.1;6.4.3.1 Main Idea;151
10.4.3.2;6.4.3.2 Example;153
10.4.3.3;6.4.3.3 Pseudocode;154
10.4.3.4;6.4.3.4 Rating;155
10.5;References;155
11;7 Interest Point Detection and Region Descriptors;156
11.1;7.1 Overview;156
11.2;7.2 Scale Invariant Feature Transform (SIFT);158
11.2.1;7.2.1 SIFT Interest Point Detector: The DoG Detector;158
11.2.1.1;7.2.1.1 Main Idea;158
11.2.1.2;7.2.1.2 Example;159
11.2.2;7.2.2 SIFT Region Descriptor;160
11.2.2.1;7.2.2.1 Main Idea;160
11.2.2.2;7.2.2.2 Example;161
11.2.3;7.2.3 Object Recognition with SIFT;161
11.2.3.1;7.2.3.1 Training Phase;161
11.2.3.2;7.2.3.2 Recognition Phase;161
11.2.3.3;7.2.3.3 Pseudocode;163
11.2.3.4;7.2.3.4 Example;164
11.2.3.5;7.2.3.5 Rating;165
11.2.3.6;7.2.3.6 Modifications;166
11.3;7.3 Variants of Interest Point Detectors;166
11.3.1;7.3.1 Harris and Hessian-Based Detectors;167
11.3.1.1;7.3.1.1 Rating;168
11.3.2;7.3.2 The FAST Detector for Corners;168
11.3.2.1;7.3.2.1 Rating;169
11.3.3;7.3.3 Maximally Stable Extremal Regions (MSER);169
11.3.3.1;7.3.3.1 Rating;170
11.3.4;7.3.4 Comparison of the Detectors;170
11.4;7.4 Variants of Region Descriptors;171
11.4.1;7.4.1 Variants of the SIFT Descriptor;171
11.4.2;7.4.2 Differential-Based Filters;173
11.4.3;7.4.3 Moment Invariants;174
11.4.4;7.4.4 Rating of the Descriptors;175
11.5;7.5 Descriptors Based on Local Shape Information;175
11.5.1;7.5.1 Shape Contexts;175
11.5.1.1;7.5.1.1 Main Idea;175
11.5.1.2;7.5.1.2 Recognition Phase;176
11.5.1.3;7.5.1.3 Pseudocode;178
11.5.1.4;7.5.1.4 Rating;179
11.5.2;7.5.2 Variants;179
11.5.2.1;7.5.2.1 Labeled Distance Sets;179
11.5.2.2;7.5.2.2 Shape Similarity Based on Contour Parts;180
11.6;7.6 Image Categorization;181
11.6.1;7.6.1 Appearance-Based ''Bag-of-Features'' Approach;181
11.6.1.1;7.6.1.1 Main Idea;181
11.6.1.2;7.6.1.2 Example;182
11.6.1.3;7.6.1.3 Modifications;183
11.6.1.4;7.6.1.4 Spatial Pyramid Matching;184
11.6.2;7.6.2 Categorization with Contour Information;185
11.6.2.1;7.6.2.1 Main Idea;186
11.6.2.2;7.6.2.2 Training Phase;187
11.6.2.3;7.6.2.3 Recognition Phase;189
11.6.2.4;7.6.2.4 Example;189
11.6.2.5;7.6.2.5 Pseudocode;190
11.6.2.6;7.6.2.6 Rating;191
11.7;References;192
12;8 Summary;194
13;Appendix A Edge Detection;198
13.1;A.1 Gradient Calculation;199
13.2;A.2 Canny Edge Detector;200
13.3;References;202
14;Appendix B Classification;203
14.1;B.1 Nearest-Neighbor Classification;203
14.2;B.2 Mahalanobis Distance;204
14.3;B.3 Linear Classification;205
14.4;B.4 Bayesian Classification;206
14.5;B.5 Other Schemes;206
14.6;References;207
15;Index;208