Imaging for Forensics and Security (eBook)

From Theory to Practice

Ahmed Bouridane (Autor)

eBook Download: PDF

2009 | 2009
XVIII, 212 Seiten
Springer US (Verlag)
978-0-387-09532-5 (ISBN)

Infos

Imaging for Forensics and Security: From Theory to Practice provides a detailed analysis of new imaging and pattern recognition techniques for the understanding and deployment of biometrics and forensic techniques as practical solutions to increase security. It contains a collection of the recent advances in the technology ranging from theory, design, and implementation to performance evaluation of biometric and forensic systems. This book also contains new methods such as the multiscale approach, directional filter bank, and wavelet maxima for the development of practical solutions to biometric problems.

The book introduces a new forensic system based on shoeprint imagery with advanced techniques for use in forensics applications. It also presents the concept of protecting the originality of biometric images stored in databases against intentional and unintentional attacks and fraud detection data in order to further increase the security.

Inhalt

Preface 7
Contents 12
1 Introduction and Preliminaries on Biometricsand Forensics Systems 17
1.1 Introduction 17
1.2 Definition of Biometrics 17
1.2.1 Biometric Characteristics 18
1.2.2 Biometric Modalities 18
1.3 Recognition/Verification/Watch-List 21
1.3.1 Verification: Am I Who I Claim to Be? 21
1.3.2 Recognition: Who Am I? 21
1.3.3 The Watch-List: Are You Looking for Me? 22
1.4 Steps of a Typical Biometric Recognition Application 22
1.4.1 Biometric Data Localisation 22
1.4.2 Normalisation and Pre-processing 23
1.4.3 Feature Extraction 24
1.4.4 Matching 25
1.4.5 Databases 25
1.5 Summary 25
References 26
2 Data Representation and Analysis 27
2.1 Introduction 27
2.2 Data Acquisition 28
2.2.1 Sensor Module 29
2.2.2 Data Storage 30
2.2.2.1 Raw Images 30
2.2.2.2 Feature Sets 30
2.3 Feature Extraction 31
2.4 Matcher 32
2.5 System Testing 32
2.6 Performance Evaluation 33
2.7 Conclusion 34
References 35
3 Improving Face Recognition Using Directional Faces 36
3.1 Introduction 36
3.2 Face Recognition Basics 37
3.2.1 Recognition/Verification 37
3.2.1.1 Face Verification: Am I Who I Claim to be? 37
3.2.1.2 Face Recognition: Who am I? 37
3.2.1.3 The Watch-List: Are You Looking for Me? 38
3.2.2 Steps of a Typical Face Recognition Application 38
3.2.2.1 Face Localisation 38
3.2.2.2 Normalisation and Pre-processing 39
3.2.2.3 Feature Extraction 40
3.2.2.4 Matching 41
3.3 Previous Work 41
3.3.1 Principal Component Analysis (PCA) 41
3.3.2 Independent Component Analysis (ICA) 42
3.3.3 Linear Discriminant Analysis (LDA) 43
3.3.4 Subspace Discriminant Analysis (SDA) 44
3.3.4.1 Dividing Classes into Subclasses 44
3.3.4.2 NN-Clustering 45
3.4 Face Recognition Using Filter Banks 46
3.4.1 Gabor Filter Bank 46
3.4.1.1 Gabor Functions and Wavelets 46
3.4.1.2 Gabor Filter Dictionary Design 47
3.4.1.3 Augmented Gabor-Face Vector 47
3.4.2 Directional Filter Bank: A Review 48
3.4.2.1 Analysis Filters 49
3.4.2.2 Quincunx Downsampling 50
3.4.2.3 Overview of --band DFB 50
3.4.2.4 Directional Images 51
3.5 Proposed Method and Results Analysis 52
3.5.1 Proposed Method 52
3.5.2 PCA 53
3.5.3 ICA 54
3.5.4 LDA 56
3.5.5 SDA 56
3.5.6 FERET Database Results 58
3.6 Conclusion 59
References 60
4 Recent Advances in Iris Recognition: A Multiscale Approach 64
4.1 Introduction 64
4.2 Related Work: A Review 66
4.3 Iris Localisation 67
4.3.1 Background 67
4.3.2 Iris Segmentation 67
4.3.3 Existing Methods for Iris Localisation 68
4.3.3.1 Daugmans Integro-Differential Operator 68
4.3.3.2 Hough Transform 69
4.3.3.3 Discrete Circular Active Contours 69
4.4 Proposed Method for Iris Localisation 70
4.4.1 Motivation 70
4.4.1.1 Edge Detector Using Wavelets 70
4.4.1.2 Multiscale Edge Detection 71
4.4.2 The Multiscale Method 72
4.4.2.1 Edge Map Detection 72
4.4.2.2 Iris Outer and Pupil Circle Detection 77
4.4.2.3 Eyelids and Eyelashes Isolating 78
4.4.2.4 Iris Normalisation and Polar Transformation 78
4.4.3 Results and Analysis 80
4.5 Texture Analysis and Feature Extraction 81
4.5.1 Wavelet Maxima Components 83
4.5.2 Special Gabor Filter Bank 83
4.5.3 Proposed Method 85
4.5.3.1 Template Generation 86
4.6 Matching 86
4.7 Experimental Results and Analysis 87
4.7.1 Database 87
4.7.2 Combined Multiresolution Feature Extraction Techniques 87
4.7.3 Template Computation 88
4.7.4 Comparison with Existing Methods 88
4.8 Discussion and Future Work 89
4.9 Conclusion 90
References 90
5 Spread Transform Watermarking Using Complex Wavelets 93
5.1 Introduction 93
5.2 Wavelet Transforms 93
5.2.1 Dual Tree Complex Wavelet Transform 94
5.2.2 Non-redundant Complex Wavelet Transform 97
5.3 Visual Models 100
5.3.1 Chou's Model 101
5.3.2 Loo's Model 107
5.3.3 Hybrid Model 108
5.4 Watermarking as Communication with Side Information 108
5.4.1 Quantisation Index Modulation 110
5.4.2 Spread Transform Watermarking 111
5.5 Proposed Algorithm 112
5.5.1 Encoding of Watermark 113
5.5.2 Decoding of Watermark 114
5.6 Information Theoretic Analysis 114
5.6.1 Decoding of Watermark 115
5.6.2 Parallel Gaussian Channels 116
5.6.3 Watermarking Game 119
5.6.4 Non-iid Data 124
5.6.5 Fixed Embedding Strategies 125
5.7 Conclusion 126
References 127
6 Protection of Fingerprint Data Using Watermarking 130
6.1 Introduction 130
6.2 Generic Watermarking System 132
6.3 State-of-the-Art 136
6.4 Optimum Watermark Detection 137
6.5 Statistical Data Modelling and Application to Watermark Detection 140
6.5.1 Laplacian and Generalised Gaussian Models 141
6.5.2 Alpha Stable Model 142
6.6 Experimental Results 143
6.6.1 Experimental Modelling of DWT Coefficients 145
6.6.2 Experimental Watermarking Results 148
6.6.2.1 Imperceptibility Analysis 148
6.6.2.2 Detection Performance 149
6.7 Conclusions 151
References 152
7 Shoemark Recognition for Forensic Science: An EmergingTechnology 155
7.1 Background to the Problem of Shoemark Forensic Evidence 155
7.1.1 Applications of a Shoemark in Forensic Science 156
7.1.2 The Need for Automating Shoemark Classification 158
7.1.3 Inconsistent Classification 159
7.1.4 Importable Classification Schema 160
7.1.5 Shoemark Processing Time Restrictions 161
7.2 Collection of Shoemarks at Crime Scenes 161
7.2.1 Shoemark Collection Procedures 162
7.2.2 Transfer/Contact Shoemarks 162
7.2.3 Photography of Shoemarks 163
7.2.4 Making Casts of Shoemarks 164
7.2.5 Gelatine Lifting of Shoemarks 165
7.2.6 Electrostatic Lifting of Shoemarks 165
7.2.7 Recovery of Shoemarks from Snow 166
7.2.8 Recovery of Shoemarks using Perfect Shoemark Scan 166
7.2.9 Making a Cast of a Shoemark Directly from a Suspect's Shoe 167
7.2.10 Processing of Shoemarks 167
7.2.11 Entering Data into a Computerised System 169
7.3 Typical Methods for Shoemark Recognition 169
7.3.1 Feature-Based Classification 170
7.3.2 Classification Based on Accidental Characteristics 171
7.4 Review of Shoemark Classfication Systems 172
7.4.1 SHOE-FIT 172
7.4.2 SHOE© 172
7.4.3 Alexandre's System 173
7.4.4 REBEZO 173
7.4.5 TREADMARK TM 174
7.4.6 SICAR 174
7.4.7 SmART 174
7.4.8 De Chazal's System 175
7.4.9 Zhang's System 175
References 175
8 Techniques for Automatic Shoeprint Classification 177
8.1 Current Approaches 177
8.2 Using Phase-Only Correlation 177
8.2.1 The POC Function 178
8.2.2 Translation and Brightness Properties of the POC Function 180
8.2.3 The Proposed Phase-Based Method 180
8.2.3.1 Spectral Weighting Function 180
8.2.3.2 Shoeprints Matching Algorithm 181
8.2.4 Experimental Results 182
8.3 Deployment of ACFs 184
8.3.1 Shoeprint Classification Using ACFs 185
8.3.2 Matching Metrics 187
8.3.3 Optimum Trade-Off Synthetic Discriminant Function Filter 188
8.3.4 Unconstrained OTSDF Filter 189
8.3.5 Tests and Results 190
8.4 Conclusion 191
References 192
9 Automatic Shoeprint Image Retrieval Using Local Features 193
9.1 Motivations 193
9.2 Local Image Features 193
9.2.1 New Local Feature Detector: Modified Harris--Laplace Detector 194
9.2.1.1 Modified Harris--Laplace (MHL) Detector 195
9.2.1.2 Repeatability Evaluation 197
9.2.2 Local Feature Descriptors 198
9.2.3 Similarity Measure 200
9.3 Experimental Results 201
9.3.1 Shoeprint Image Databases 201
9.4 Summary 210
References 212
Index 214

Produkt-Details

• Erscheint lt. Verlag: 14.7.2009
• Reihe/Serie: Signals and Communication Technology
• Reihe/Serie: Signals and Communication Technology
• Zusatzinfo: XVIII, 212 p. 60 illus.
• Verlagsort: New York
• Sprache: englisch
• Themenwelt: Informatik ► Grafik / Design ► Digitale Bildverarbeitung
• Themenwelt: Informatik ► Netzwerke ► Sicherheit / Firewall
• Themenwelt: Informatik ► Theorie / Studium ► Künstliche Intelligenz / Robotik
• Themenwelt: Medizin / Pharmazie ► Gesundheitsfachberufe
• Themenwelt: Studium ► 2. Studienabschnitt (Klinik) ► Rechtsmedizin
• Themenwelt: Technik ► Elektrotechnik / Energietechnik
• Schlagworte: biometrics • classification • computer vision • fingerprint • Forensic Imagery • pattern recognition • Shoeprint Recognition
• ISBN-10: 0-387-09532-2 / 0387095322
• ISBN-13: 978-0-387-09532-5 / 9780387095325