Title Radar automatic target recognition (ATR) and non-cooperative target recognition (NCTR) / edited by David Blacknell, Hugh Griffiths

Published London, United Kingdom : The Institute of Engineering and Technology, 2013

Table of Contents
1.	Introduction					1
1.1.	Motivation					1
1.2.	Definitions and acronyms					2
1.3.	Scope of book					3
2.	Automatic target recognition of ground targets					5
2.1.	Introduction					5
2.2.	SAR phenomenology					7
2.3.	The ATR processing chain					11
2.3.1.	Pre-screening					11
2.3.2.	Template-matching					14
2.3.3.	Feature-based classification					15
2.4.	Use of contextual information in target detection					19
2.4.1.	Motivation					19
2.4.2.	Statistical formulation					19
2.4.3.	Simulated results					21
2.5.	Databases and modelling					22
2.5.1.	Database construction					22
2.5.2.	Case study: model-based ATR using MOCEM					24
2.6.	Performance assessment					26
2.6.1.	Receiver operating characteristic (ROC) curves					26
2.6.2.	Confusion matrices					29
2.6.3.	Operational assessment					32
2.7.	Conclusions					34
	Acknowledgements					34
	References					35
3.	Automatic recognition of air targets					37
3.1.	Introduction					37
3.2.	Fundamentals of the target recognition process					38
3.2.1.	Introduction					38
3.2.2.	Target features					38
3.2.3.	Aircraft recognition techniques and waveform design					39
3.2.4.	Target signature measurement					41
3.2.5.	Radar range equation for radar target recognition					42
3.2.6.	Main classification functions					43
3.2.7.	Database					44
3.2.8.	Classifier					44
3.2.9.	Assembly of database					46
3.2.10.	Classifier performance					47
3.2.11.	Conclusions					48
3.3.	Jet engine recognition					48
3.3.1.	Introduction					48
3.3.2.	Jet engine mechanics					49
3.3.3.	Interaction of radar signal with engine blades					49
3.3.4.	Jet engine modulation spectrum: engine rotational rate					50
3.3.5.	Jet engine modulation spectrum: rotor stage spectrum					52
3.3.6.	Jet engine modulation spectrum: mixing products from rotor stages					54
3.3.7.	Determination of blade count					55
3.3.8.	JEM waveform					55
3.3.9.	System requirements					56
3.3.10.	Conclusions					56
3.4.	Helicopter recognition					57
3.4.1.	Introduction					57
3.4.2.	Main rotor blade flash					57
3.4.3.	Detection of blade flash					60
3.4.4.	Waveform and system requirements for blade flash detection					62
3.4.5.	Blade flash detection					62
3.4.6.	Helicopter classification using blade flash					63
3.4.7.	Main rotor hub spectrum					63
3.4.8.	Rear rotor blades					65
3.4.9.	Radar range equation for helicopter recognition					66
3.4.10.	Helicopter recognition summary					67
3.5.	Range-Doppler imaging					67
3.5.1.	Introduction					67
3.5.2.	Helicopter signature					69
3.5.3.	Jet airliner signature					70
3.5.4.	Business jet signature					71
3.5.5.	Propeller aircraft signature					72
3.5.6.	Waveforms and system requirements for supporting RDI					72
3.5.7.	Conclusions					73
3.6.	Aircraft target recognition conclusions					73
	Acknowledgements					74
	References					74
4.	Radar ATR of maritime targets					77
4.1.	Introduction					77
4.2.	The use of high range resolution (HRR) profiles for ATR					78
4.3.	The derivation of ATR features from HRR profiles					80
4.3.1.	Length estimate					80
4.3.2.	Position specific matrices (PSMs)					83
4.3.2.1.	Determination of length					83
4.3.2.2.	Alignment					83
4.3.2.3.	Quantisation					83
4.3.2.4.	Creation of reference PSMs					84
4.3.2.5.	Compare the quantised test profile to the reference PSMs					84
4.3.2.6.	Determine a figure of merit					84
4.3.2.7.	Classification					86
4.3.3.	Other examples of ATR features					86
4.3.4.	Choosing sets of uncorrelated features					87
4.4.	Ship ATR under the influence of multipath					88
4.4.1.	What is multipath?					88
4.4.2.	The problem of defining testing and training vectors					90
4.5.	Results					92
4.5.1.	Length estimate					92
4.5.1.1.	Results for La and Lb based on measurements of ship HRR profiles					92
4.5.1.2.	Simulation of ship HRR profiles					94
4.5.2.	PSM results					96
4.5.3.	Results based on geometrical, statistical and structural features					99
4.5.3.1.	Measurements					99
4.5.3.2.	Classification based on simulated ships					104
4.6.	The mitigation of multipath effects on ship ATR					107
4.6.1.	Using several antennas					109
4.6.2.	Using several frequencies					110
4.6.3.	Combining two antennas and two frequencies					114
4.6.4.	Classification improvement via multi-frequency and/or multi-antenna approach					120
4.7.	Summary					123
	References					125
5.	Effects of image quality on target recognition					127
5.1.	Introduction					127
5.2.	Improving ATR performance via PGA image quality enhancement					128
5.3.	Improving ATR performance using high resolution, PWF-processed full-polarisation SAR data					131
5.4.	Improving ATR performance via high-definition image processing					138
5.5.	Reconstruction of interrupted SAR imagery					147
5.6.	Summary and conclusions					153
	References					153
6.	Comparing classifier effectiveness					157
6.1.	Introduction					157
6.2.	NCTI studies					158
6.3.	Measurements					158
6.3.1.	TIRA system					158
6.3.2.	Targets					160
6.4.	Idea of classification					160
6.4.1.	Appropriate features					160
6.4.2.	HRR and 2D ISAR					161
6.4.3.	2D ISAR template correlation classifier					164
6.4.4.	Selection of radar parameters					166
6.5.	Classification scheme					166
6.5.1.	Pre-processing unit					167
6.5.2.	Feature extraction/reduction					168
6.5.3.	Choosing a classifier					169
6.5.4.	Test of classifiers					170
6.6.	Feature extraction					171
6.6.1.	Classification results using different feature sets					172
6.7.	Conclusion					174
	References					174
7.	Biologically inspired and multi-perspective target recognition					177
7.1.	Introduction					177
7.2.	Biologically inspired NCTR					179
7.2.1.	Waveform design					179
7.2.2.	Nectar-feeding bats and bat-pollinated plants					180
7.2.3.	Classification of flowers					181
7.2.3.1.	Data collection					182
7.2.3.2.	Data pre-processing and results					182
7.2.4.	Classification of insects					186
7.3.	Acoustic micro-Doppler					188
7.3.1.	Description of the acoustic radar					190
7.3.2.	Experimentation					190
7.3.3.	Classification performance results					193
7.4.	Multi-aspect NCTR					194
7.4.1.	Data preparation					199
7.4.2.	Feature extraction					199
7.4.3.	Multi-perspective classifiers					200
7.4.4.	Multi-perspective classification performance					202
7.5.	Summary					206
	References					208
8.	Radar applications of compressive sensing					213
8.1.	Introduction					213
8.2.	Principles of compressive sensing					214
8.2.1.	Sparse and compressible signals					214
8.2.2.	Restricted isometric property and coherence					216
8.2.3.	Signal reconstruction					217
8.2.3.1.	Minimum l2 norm reconstruction					218
8.2.3.2.	Minimum l0 norm reconstruction					218
8.2.3.3.	Minimum l1 norm reconstruction					218
8.2.3.4.	Example of l1 norm versus l2 norm reconstruction					219
8.3.	Reconstruction algorithms					220
8.3.1.	Convex optimisation					220
8.3.1.1.	Basis pursuit					220
8.3.1.2.	Basis pursuit de-noising					221
8.3.1.3.	Least absolute shrinkage and selection operator					221
8.3.2.	Greedy constructive algorithms					222
8.3.2.1.	Matching pursuit					222
8.3.2.2.	Orthogonal matching pursuit					223
8.3.2.3.	Stage-wise orthogonal matching pursuit					223
8.3.3.	Iterative thresholding algorithms					224
8.3.3.1.	Iterative hard thresholding					225
8.3.3.2.	Iterative shrinkage and thresholding					226
8.4.	Jet engine modulation					226
8.4.1.	Introduction					226
8.4.2.	Jet engine model					227
8.4.3.	Simulation results of JEM compressive sensing					228
8.5.	Inverse synthetic aperture radar					230
8.5.1.	Introduction					230
8.5.2.	Simulation model					231
8.6.	Conclusions					234
	Acknowledgements					234
	References					234
9.	Advances in SAR change detection					237
9.1.	Introduction					237
9.2.	An analysis of the CCD algorithm					239
9.3.	Results using the ùniversal image quality index'					242
9.4.	Performance comparison of change detection algorithms					245
9.4.1.	Visual comparisons of the MLE and CCD algorithms					253
9.4.2.	Coherent change detection performance with shadow regions masked					258
9.5.	Summary and conclusions					263
	References					263
10.	Future challenges					265
10.1.	Introduction					265
10.2.	Future challenges					266
10.2.1.	Target variability and practical databases					266
10.2.2.	Complex clutter environments					267
10.2.3.	Use of contextual information					268
10.2.4.	Performance assessment and prediction					269
10.2.5.	Deception and countermeasures					271
	References					271
	Index					273

Click on the following:
IET eBooks

Copies

Description 1 online resource (xii, 279 pages) Series IET radar, sonar and navigation series ; 33 IET radar, sonar, navigation and avionics series ; 33. Contents Machine generated contents note: 1. Introduction -- 1.1. Motivation -- 1.2. Definitions and acronyms -- 1.3. Scope of book -- 2. Automatic target recognition of ground targets -- 2.1. Introduction -- 2.2. SAR phenomenology -- 2.3. ATR processing chain -- 2.3.1. Pre-screening -- 2.3.2. Template-matching -- 2.3.3. Feature-based classification -- 2.4. Use of contextual information in target detection -- 2.4.1. Motivation -- 2.4.2. Statistical formulation -- 2.4.3. Simulated results -- 2.5. Databases and modelling -- 2.5.1. Database construction -- 2.5.2. Case study: model-based ATR using MOCEM -- 2.6. Performance assessment -- 2.6.1. Receiver operating characteristic (ROC) curves -- 2.6.2. Confusion matrices -- 2.6.3. Operational assessment -- 2.7. Conclusions -- Acknowledgements -- References -- 3. Automatic recognition of air targets -- 3.1. Introduction -- 3.2. Fundamentals of the target recognition process -- 3.2.1. Introduction -- 3.2.2. Target features -- 3.2.3. Aircraft recognition techniques and waveform design -- 3.2.4. Target signature measurement -- 3.2.5. Radar range equation for radar target recognition -- 3.2.6. Main classification functions -- 3.2.7. Database -- 3.2.8. Classifier -- 3.2.9. Assembly of database -- 3.2.10. Classifier performance -- 3.2.11. Conclusions -- 3.3. Jet engine recognition -- 3.3.1. Introduction -- 3.3.2. Jet engine mechanics -- 3.3.3. Interaction of radar signal with engine blades -- 3.3.4. Jet engine modulation spectrum: engine rotational rate -- 3.3.5. Jet engine modulation spectrum: rotor stage spectrum -- 3.3.6. Jet engine modulation spectrum: mixing products from rotor stages -- 3.3.7. Determination of blade count -- 3.3.8. JEM waveform -- 3.3.9. System requirements -- 3.3.10. Conclusions -- 3.4. Helicopter recognition -- 3.4.1. Introduction -- 3.4.2. Main rotor blade flash -- 3.4.3. Detection of blade flash -- 3.4.4. Waveform and system requirements for blade flash detection -- 3.4.5. Blade flash detection -- 3.4.6. Helicopter classification using blade flash -- 3.4.7. Main rotor hub spectrum -- 3.4.8. Rear rotor blades -- 3.4.9. Radar range equation for helicopter recognition -- 3.4.10. Helicopter recognition summary -- 3.5. Range-Doppler imaging -- 3.5.1. Introduction -- 3.5.2. Helicopter signature -- 3.5.3. Jet airliner signature -- 3.5.4. Business jet signature -- 3.5.5. Propeller aircraft signature -- 3.5.6. Waveforms and system requirements for supporting RDI -- 3.5.7. Conclusions -- 3.6. Aircraft target recognition conclusions -- Acknowledgements -- References -- 4. Radar ATR of maritime targets -- 4.1. Introduction -- 4.2. use of high range resolution (HRR) profiles for ATR -- 4.3. derivation of ATR features from HRR profiles -- 4.3.1. Length estimate -- 4.3.2. Position specific matrices (PSMs) -- 4.3.2.1. Determination of length -- 4.3.2.2. Alignment -- 4.3.2.3. Quantisation -- 4.3.2.4. Creation of reference PSMs -- 4.3.2.5. Compare the quantised test profile to the reference PSMs -- 4.3.2.6. Determine a figure of merit -- 4.3.2.7. Classification -- 4.3.3. Other examples of ATR features -- 4.3.4. Choosing sets of uncorrelated features -- 4.4. Ship ATR under the influence of multipath -- 4.4.1. What is multipath? -- 4.4.2. problem of defining testing and training vectors -- 4.5. Results -- 4.5.1. Length estimate -- 4.5.1.1. Results for La and Lb based on measurements of ship HRR profiles -- 4.5.1.2. Simulation of ship HRR profiles -- 4.5.2. PSM results -- 4.5.3. Results based on geometrical, statistical and structural features -- 4.5.3.1. Measurements -- 4.5.3.2. Classification based on simulated ships -- 4.6. mitigation of multipath effects on ship ATR -- 4.6.1. Using several antennas -- 4.6.2. Using several frequencies -- 4.6.3. Combining two antennas and two frequencies -- 4.6.4. Classification improvement via multi-frequency and/or multi-antenna approach -- 4.7. Summary -- References -- 5. Effects of image quality on target recognition -- 5.1. Introduction -- 5.2. Improving ATR performance via PGA image quality enhancement -- 5.3. Improving ATR performance using high resolution, PWF-processed full-polarisation SAR data -- 5.4. Improving ATR performance via high-definition image processing -- 5.5. Reconstruction of interrupted SAR imagery -- 5.6. Summary and conclusions -- References -- 6. Comparing classifier effectiveness -- 6.1. Introduction -- 6.2. NCTI studies -- 6.3. Measurements -- 6.3.1. TIRA system -- 6.3.2. Targets -- 6.4. Idea of classification -- 6.4.1. Appropriate features -- 6.4.2. HRR and 2D ISAR -- 6.4.3. 2D ISAR template correlation classifier -- 6.4.4. Selection of radar parameters -- 6.5. Classification scheme -- 6.5.1. Pre-processing unit -- 6.5.2. Feature extraction/reduction -- 6.5.3. Choosing a classifier -- 6.5.4. Test of classifiers -- 6.6. Feature extraction -- 6.6.1. Classification results using different feature sets -- 6.7. Conclusion -- References -- 7. Biologically inspired and multi-perspective target recognition -- 7.1. Introduction -- 7.2. Biologically inspired NCTR -- 7.2.1. Waveform design -- 7.2.2. Nectar-feeding bats and bat-pollinated plants -- 7.2.3. Classification of flowers -- 7.2.3.1. Data collection -- 7.2.3.2. Data pre-processing and results -- 7.2.4. Classification of insects -- 7.3. Acoustic micro-Doppler -- 7.3.1. Description of the acoustic radar -- 7.3.2. Experimentation -- 7.3.3. Classification performance results -- 7.4. Multi-aspect NCTR -- 7.4.1. Data preparation -- 7.4.2. Feature extraction -- 7.4.3. Multi-perspective classifiers -- 7.4.4. Multi-perspective classification performance -- 7.5. Summary -- References -- 8. Radar applications of compressive sensing -- 8.1. Introduction -- 8.2. Principles of compressive sensing -- 8.2.1. Sparse and compressible signals -- 8.2.2. Restricted isometric property and coherence -- 8.2.3. Signal reconstruction -- 8.2.3.1. Minimum l2 norm reconstruction -- 8.2.3.2. Minimum l0 norm reconstruction -- 8.2.3.3. Minimum l1 norm reconstruction -- 8.2.3.4. Example of l1 norm versus l2 norm reconstruction -- 8.3. Reconstruction algorithms -- 8.3.1. Convex optimisation -- 8.3.1.1. Basis pursuit -- 8.3.1.2. Basis pursuit de-noising -- 8.3.1.3. Least absolute shrinkage and selection operator -- 8.3.2. Greedy constructive algorithms -- 8.3.2.1. Matching pursuit -- 8.3.2.2. Orthogonal matching pursuit -- 8.3.2.3. Stage-wise orthogonal matching pursuit -- 8.3.3. Iterative thresholding algorithms -- 8.3.3.1. Iterative hard thresholding -- 8.3.3.2. Iterative shrinkage and thresholding -- 8.4. Jet engine modulation -- 8.4.1. Introduction -- 8.4.2. Jet engine model -- 8.4.3. Simulation results of JEM compressive sensing -- 8.5. Inverse synthetic aperture radar -- 8.5.1. Introduction -- 8.5.2. Simulation model -- 8.6. Conclusions -- Acknowledgements -- References -- 9. Advances in SAR change detection -- 9.1. Introduction -- 9.2. analysis of the CCD algorithm -- 9.3. Results using the ùniversal image quality index' -- 9.4. Performance comparison of change detection algorithms -- 9.4.1. Visual comparisons of the MLE and CCD algorithms -- 9.4.2. Coherent change detection performance with shadow regions masked -- 9.5. Summary and conclusions -- References -- 10. Future challenges -- 10.1. Introduction -- 10.2. Future challenges -- 10.2.1. Target variability and practical databases -- 10.2.2. Complex clutter environments -- 10.2.3. Use of contextual information -- 10.2.4. Performance assessment and prediction -- 10.2.5. Deception and countermeasures -- References Summary The ability to detect and locate targets by day or night, over wide areas, regardless of weather conditions has long made radar a key sensor in many military and civil applications. However, the ability to automatically and reliably distinguish different targets represents a difficult challenge. Radar Automatic Target Recognition (ATR) and Non-Cooperative Target Recognition (NCTR) captures material presented in the NATO SET-172 lecture series to provide an overview of the state-of-the-art and continuing challenges of radar target recognition. Topics covered include the problem as applied to the ground, air and maritime domains; the impact of image quality on the overall target recognition performance; the performance of different approaches to the classifier algorithm; the improvement in performance to be gained when a target can be viewed from more than one perspective; the impact of compressive sensing; advances in change detection; and challenges and directions for future research. Radar Automatic Target Recognition (ATR) and Non-Cooperative Target Recognition (NCTR) explores both the fundamentals of classification techniques applied to data from a variety of radar modes and selected advanced techniques at the forefront of research, and is essential reading for academic, industrial and military radar researchers, students and engineers worldwide Analysis military radar researchers change detection compressive sensing classifier algorithm target recognition performance image quality NATO SET-172 lecture series key sensor NCTR noncooperative target recognition ATR radar automatic target recognition Bibliography Includes bibliographical references and index Subject Radar. Radar targets. Radar -- Military applications. Signal processing. Target acquisition. Radar radar. Radar targets Radar Radar -- Military applications Signal processing Target acquisition compressed sensing. military radar. object detection. radar target recognition. Form Electronic book Author Blacknell, David, editor. Griffiths, H. (Hugh), 1956- editor. ISBN 1849196869 9781849196864

  Permalink