Physics of Automatic Target Recognition (eBook)

Contents 7
Preface 9
Contributors 10
Kernel-Based Nonlinear Subspace Target Detection for Hyperspectral Imagery 12
1.1 Introduction 12
1.2 Kernel Methods and Kernel Trick 13
1.3 Linear Matched Subspace Detector and Kernel Matched Subspace Detector 14
1.4 OSP and Kernel OSP Algorithms 18
1.5 Linear Spectral Matched Filter and Kernel Spectral Matched Filter 21
1.6 Adaptive Subspace Detector and Kernel Adaptive Subspace Detector 23
1.7 Experimental Results 25
1.8 Conclusions 29
1.9 Appendix I (Kernel PCA) 31
References 34
Theory of Invariant Algebra and Its Use in Automatic Target Recognition 36
2.1 Introduction 36
2.2 Theory of Invariant Alagebra 37
2.3 Applications of Invariants of Binary Quantics 39
2.4 Invariant of Ternary Quantics 43
2.5 Complexity Analysis 46
2.6 Summary 50
References 51
Automatic Recognition of Underground Targets Using Time– Frequency Analysis and Optimization Techniques 52
3.1 Introduction 52
3.2 Theoretical Models 54
3.3 Classification of Test Field Targets 59
3.4 Summary and Conclusions 65
Acknowledgements 66
References 66
A Weighted Zak Transform, Its Properties, and Applications to Signal Processing 68
4.1 Introduction 68
4.2 Background and Definitions 69
4.3 The Weighted Zak Transform and the Ambiguity Function 70
4.4 The Weighting Function 72
4.5 The Inverse of the Weighted Zak Transform 75
4.6 Signal Synthesis Using the Weighted Zak Transform 81
Acknowledgment 82
References 82
Using Polarization Features of Visible Light for Automatic Landmine Detection 83
5.1 Introduction 83
5.2 Landmine Detection with Camera Systems 83
5.3 Detection 85
5.4 Detection Tests 88
5.5 Test Results 93
5.6 Conclusions 98
References 99
The Physics of Polarization-Sensitive Optical Imaging 101
6.1 Introduction 101
6.2 Stokes Vectors and Mueller Matrices 102
6.3 Target Shape Information from Polarization Imaging 102
6.4 Summary 113
References 113
Dispersion, Its Effects, and Compensation 115
7.1 Introduction 115
7.2 Basic Ideas and Definitions 117
7.3 Spatial Moments of a Pulse 118
7.4 Dispersion-Invariant Moments 122
7.5 Example 124
7.6 Wigner Approximation for Wave Propagation, and Approximately Invariant Moments 128
7.7 Dual Case: Initially Given u(0, t) 130
7.8 Conclusion 134
Acknowledgment 135
References 135
Multisensor Target Recognition in Image Response Space Using Evolutionary Algorithms 136
8.1 Introduction 136
8.2 Hybrid Model of Evolutionary Algorithm in Target Recognition with Different Imagery Types 138
8.3 Computational Experiments with Different Imagery Types 141
8.4 Conclusions 148
Acknowledgement 149
References 149
Biophysics of the Eye in Computer Vision: Methods and Advanced Technologies 151
9.1 Chapter Overview 151
9.2 Background Review: Light, Eye Responses, and Movements 153
9.3 Eye Appearance Changes 156
9.4 Eye Image Representation 158
9.5 Modeling of the Eye Variability 159
9.6 Automatic Eye Localization 161
9.7 Tracking Problem Formulation 164
9.8 Models of Tracking Methods 166
9.9 Particle Filter-Based Eye Tracking Methods 167
9.10 Advanced Eye Tracking–Based Technologies 176
9.11 Conclusion 178
References 178
Two Approaches to 3D Microorganism Recognition Using Single Exposure Online ( SEOL) Digital Holography 183
10.1 Introduction 183
10.2 Single Exposure Online (SEOL) Digital Holography 185
10.3 3D Morphology-Based Recognition 187
10.4 Shape-Tolerant 3D Recognition 189
10.5 Experimental Results 193
10.6 Conclusion 200
References 200
Distortion-Tolerant 3D Object Recognition by Using Single Exposure On- Axis Digital Holography 203
11.1 Introduction 203
11.2 Single Exposure On-Axis Digital Holography 204
11.3 Composite Filter Design 206
11.4 Experimental Results 207
11.5 Conclusion 213
References 214
Design of Distortion-Invariant Optical ID Tags for Remote Identification and Verification of Objects 215
12.1 Introduction 215
12.2 Authentication System 216
12.3 Signature Encryption and Decryption 217
12.4 Distortion-Invariant ID Tags 219
12.5 Signature Verification 224
12.6 Authentication Results 225
12.7 Conclusions 232
References 232
Speckle Elimination With a Maximum Likelihood Estimation and an Isoline Regularization 235
13.1 Presentation 235
13.2 Description of the Algorithm 236
13.3 Results and Influence of the Parameters 243
13.4 Conclusion 249
Acknowledgment 250
References 250
Index 252