Title Page
Contents
ABSTRACT 7
Chapter 1. Introduction 9
1.1. Introduction 9
Chapter 2. LiDAR-vehicle Calibration 13
2.1. Introduction to LiDAR-vehicle Calibration 13
2.2. Feature-based Method 14
2.3. Motion-based Method 16
Chapter 3. Online LiDAR-vehicle Calibration Framework 20
3.1. Z, Roll, and Pitch (zϕθ) Estimation 21
3.2. X, Y, and Yaw (xyΨ) Estimation 23
Chapter 4. Fault Data Exclusion 24
Chapter 5. Observable Data Balancing 27
Chapter 6. Experiments 29
6.1. Experiments Environments 29
6.2. Online Calibration Performance 31
6.3. Ablation Experiments 33
6.3.1. Ablation of Fault Data Filtering 33
6.3.2. Ablation of Observable Data Balancing 35
Chapter 7. Conclusion 37
References 38
Abstract (in Korean) 43
〈Figure 1-1〉 Autonomous vehicle application 10
〈Figure 2-1〉 LiDAR-vehicle extrinsic parameter 13
〈Figure 2-2〉 Different measurement from LiDAR and INS/GNSS 13
〈Figure 2-3〉 Plane feature-based Calibration (Before optimization) 15
〈Figure 2-4〉 Plane feature-based Calibration (After optimization) 15
〈Figure 2-5〉 LiDAR odometry and vehicle odometry 16
〈Figure 3-1〉 System architecture 20
〈Figure 3-2〉 Ground extraction and uneven ground exclusion 22
〈Figure 4-1〉 Registration fault 24
〈Figure 5-1〉 Curvature-based data balancing 28
〈Figure 5-2〉 Motion-based calibration using balanced data 28
〈Figure 6-1〉 Autonomous vehicle with INS/GNSS sensor and LiDAR sensors 29
〈Figure 6-2〉 City road scenario and highway scenario 30
〈Figure 6-3〉 zϕθ disturbance experiments results 31
〈Figure 6-4〉 xyΨ disturbance experiments results 32
〈Figure 6-5〉 Ablation of fault data filtering 34
〈Figure 6-6〉 Ablation of observable data balancing 36