0%

LIO-SAM

1. What

Lidar SLAM coupled with IMU and GPS (optional)

2. Components

LIO-SAM 基本流程

  • ImuPreintergration
  • ImageProjeection
  • Feature Extraction
  • MapOptimization

Two odometry

  • Incremental, use only local matching
  • Global, for loop detection and optimization

3. IMU

  • Input: raw IMU, lidar odometry output
  • Output: local IMU odometry, global IMU odometry

IMUPreintegration 类

  • imu: used to directly calcualte the pose
  • imuOpt: used to optimization

4. ImageProjection

  • Input: raw IMU, IMU odometry output, and raw point cloud
  • Output: corrected point cloud with related index

5. Feature Extraction

Almost the same as LOAM

  • corner point
  • plane point

6. MapOptimization

Input: Point Cloud, GPS

Output: lidar odometry (global and incremental)

  • Main thread
  • loop detection loop