2D Pose Graph Optimization

Introduction

This is a C++ implementation of an efficient 2D pose graph optimization, by exploiting the sparsity in the graph structure, as outlined in [1, 2]

Installation

Dependencies

Eigen, for optimization.
Qt Creator, optional, if you want to run the example.

sudo apt-get update
sudo apt-get install qtcreator

Source Code

The source code is hosted on GitHub: ShahramKhorshidi/pose_graph_optimization

# Clone pose_graph_optimization
git clone [email protected]:ShahramKhorshidi/pose_graph_optimization.git

Build

cd pose_graph_optimization
mkdir build && cd build/ 
cmake ..
make

Quick Start

You can run the example with recorded trajectory of a wheeled robot. Before running the example change the main.cpp at line 24, adapting the path to your local repository.

cd pose_graph_optimization/build/example
./RobotVisualization

Framework

References

[1] G. Grisetti, R. Kümmerle, C. Stachniss and W. Burgard, "A Tutorial on Graph-Based SLAM," in IEEE Intelligent Transportation Systems Magazine, 2010.

[2] K. Konolige, G. Grisetti, R. Kümmerle, W. Burgard, B. Limketkai and R. Vincent, "Efficient Sparse Pose Adjustment for 2D mapping," 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2010.