Download
SLAM in Autonomous Driving Book is the open-source code companion to the book SLAM in Autonomous Driving. It focuses on localization and mapping systems that combine LiDAR, IMU, GNSS, and vehicle motion data. The material begins with geometry, kinematics, Kalman filtering, inertial navigation, preintegration, and graph optimization. It then implements point-cloud processing, nearest-neighbor structures, ICP, NDT, scan matching, occupancy grids, and loop closure. Readers build both loosely coupled and tightly coupled LiDAR-inertial odometry systems. Later modules cover offline mapping, backend optimization, tiled map export, map loading, initialization, and fused real-time localization. The repository includes configuration files, Docker support, demonstrations, and datasets for urban, campus, road, mall, and solid-state LiDAR scenarios.
Features
- IMU, GNSS, and LiDAR sensor fusion
- Error-state Kalman filter implementations
- ICP, NDT, and scan-matching algorithms
- Two-dimensional and three-dimensional LiDAR mapping
- Loosely and tightly coupled LIO systems
- Offline mapping and fused localization
Project Samples
