Graph-Based SLAM
Graph-based SLAM is a method of representing the environment and the robot’s state using a graph.
- graph encode the geometric constraints
- We then do Pose Graph Optimization to estimate the robot’s trajectory and the map
So it doesn’t do Scan Matching, unlike EKF-SLAM.
These heavily rely on the Loop Closure to work well.
http://ais.informatik.uni-freiburg.de/teaching/ws11/robotics2/pdfs/ls-slam-tutorial.pdf
Pose Graph Optimization
This is the optimization technique used