Skip to content

Chapter 4: Deployment of SLAM

4.1 Objectives

4.2 Available datasets

4.3 What is a high-quality input

4.4 Available SLAM approaches

4.4.1 Optimization-based (one example for each)

4.4.2 Filter-based

4.4.3 Loosely coupled

4.4.4 Tightly coupled

4.4.5 Deep learning-based

4.5 Deploying a camera-IMU SLAM

4.5.1 A camera-IMU odom method using a SMARTNav bag file

4.5.2 Odom + loop closure

4.5.3 Parameter tuning

4.6 Deploying a LiDAR-IMU SLAM

4.6.1 Running FASTLIO

4.6.2 Running GLIM or Lego-LOAM

4.6.3 Parameters

4.7 Practical consideration

4.7.1 Processing unit

4.7.2 Standalone or ROS-based

4.7.3 QoS in ROS

4.7.4 Being real-time and low delay

4.7.5 Using in control loops