Sensor Fusion Approaches for Positioning, Navigation, and Mapping

About the Author xix

Preface xxi

Acknowledgment xxvii

1 Coordinate Systems and Motion Modeling 1

1.1 Introduction 1

1.2 Points, Vectors, and Matrices 1

1.3 Coordinate Frames and Systems 12

1.4 Transformation Between Frames 19

1.5 The Mathematics of Rigid Body Motion in 3D Space 26

1.6 Problems 40

References 43

2 Signals and Systems 45

2.1 Introduction 45

2.2 Probability 45

2.3 Systems 55

2.4 Signals 71

2.5 Random Processes 76

2.6 Signal Processing Basics 82

2.7 Problems 100

References 102

3 Sensor Fusion Methods and Algorithms 103

3.1 Introduction 103

3.2 Estimation Philosophy 103

3.3 Gauss-Markov Process Model 104

3.4 State Estimation 105

3.5 Machine Learning and Artificial Intelligence 131

3.6 Multisensor Temporal and Spatial Extrinsic Calibration 164

3.7 Problems 165

References 166

4 Inertial Sensors and Inertial Navigation Systems 169

4.1 Introduction 169

4.2 Inertial Measurements 169

4.3 Earth Gravity/Rotation/Curvature Effects on Inertial Measurements 174

4.4 Inertial Measurement Unit 178

4.5 Inertial Sensor Errors 180

4.6 Inertial Navigation Systems (INS) 188

4.7 Problems 201

References 202

5 Radio Positioning Systems 203

5.1 Introduction 203

5.2 Concept of Operation 203

5.3 The Electromagnetic Waves and Radio Spectrum 205

5.4 Radio Positioning Transmitter 206

5.5 Positioning Signal Structure 20

5.6 Receiver Signal Processing 214

5.7 Radio Positioning Schemes 228

5.8 Dilution of Precision (DOP) 234

5.9 GNSS Technology 235

5.10 GNSS Positioning Methods 24

5.11 GPS SPS MATLAB Project Sample 259

5.12 Sagnac Effect 261

5.13 Problems 264

References 265

6 Active Ranging Sensors 267

6.1 Introduction 267

6.2 Principle of Operation of Active Ranging/Detection 267

6.3 Sound, Radio, and Light Detection and Ranging 273

6.4 Signal Processing of Ranging Sensors 275

6.5 MATLAB Implementation of FMCW Ranging Signal Processing 289

6.6 Range-Based Odometry 290

6.7 Problems 301

References 302

7 Imaging Sensors 303

7.1 Introduction 303

7.2 How Imaging Sensors Work 303

7.3 Camera Projection Model 304

7.4 Estimating Camera Motion from Images 308

7.5 Visual 3D Reconstruction: Estimating Physical Points’ Location 318

7.6 Optical Flow Estimation 319

7.7 Stereo Camera System 32

7.8 Camera Calibration 325

7.9 The Fundamental Matrix 327

7.10 Enhanced Motion Estimation from More Than Two Images 328

7.11 Visual Odometry 331

7.12 Problems 338

References 339

8 Mapping Algorithms 341

8.1 Introduction 341

8.2 Occupancy Grid Mapping 342

8.3 MATLAB Occupancy Grid Building Example 346

8.4 SLAM Problem and Solution Approaches 348

8.5 Visual SLAM 363

References 373

9 Case Study #1: Wheeled Platforms 375

9.1 Introduction 375

9.2 Reference Frames and the Lever Arm 376

9.3 2D System Model 378

9.4 2D Error Model 380

9.5 Loosely Coupled GPS Measurement Model 382

9.6 Tightly Coupled GPS Measurement Model 386

9.7 GPS Fusion Using Extended Kalman Filter 391

9.8 GPS Fusion Using Particle Filter (PF) 393

9.9 GPS Deeply Coupled Fusion 394

9.10 Fusing Wheel Encoder Data 395

9.11 Fusing LiDAR with 2D Inertial/Wheel Encoder Sensors 395

9.12 Graph SLAM Fusion of Inertial, Wheel-Odometry, and LiDAR 401

9.13 MATLAB Project Samples for 2D Wheeled Platforms 403

10 Case Study #2: Aerial Vehicles 415

10.1 Introduction 415

10.2 3D Motion System Model 415

10.3 Fusion of GPS Measurements 424

10.4 Fusion of Vision Measurements 426

10.5 Fusion of RADAR Measurements 437

10.6 MATLAB Implementation of 3D IMU/GNSS Fusion Using EKF 441

11 Case Study #3: AHRS and PDR 457

11.1 Introduction 457

11.2 Attitude and Heading Reference System (AHRS) 458

11.3 Pedestrian Dead Reckoning (PDR) 469

11.4 MATLAB Project Sample of an AHRS Using EKF 473

12 Learning-Based Fusion Methods 481

12.1 Introduction 481

12.2 Learning-Based Approaches 482

12.3 Formulation of Loss Function 484

12.4 Recent Trends and Examples 488

References 500

Index 503