Toy Pose Graph Optimization with CERES

Web URL : https://kusemanohar.wordpress.com/2017/04/29/howto-pose-graph-bundle-adjustment/.

Author : Manohar Kuse : mpkuse [At] connect [.] ust [dot] hk

Requirements

You need to install ceres-solver and Eigen3 before you can compile this code.

How to compile

Compile

mkdir build
cd build
cmake ..
make

This should produce an executable toy_pose_graph. Run this executable from build folder.

Run Executable

./toy_pose_graph

This executable reads file ../input_M3500_g2o.g2o and produces ../init_nodes.txt, ../init_edges.txt ../after_opt_nodes.txt, ../after_opt_edges.txt and ../switches.txt

Visualize Results

We have provided a python script to visualize the results. The text files to supply should contain lines as : id x y theta representing every node.

cd .. #come out of build directory
python plot_results.py --initial_poses init_nodes.txt --optimized_poses after_opt_nodes.txt

List of Files

• ceres_try.cpp --> Code to read .g2o file and define cost function
• input_M3500_g2o.g2o --> Sample pose-graph. The manhattan dataset. More benchmarking pose-graph datasets.
• plot_results.py --> Python script to visualize the results

Result