pip install evo --upgrade --no-binary evo
mkdir build
cd build
cmake ..
make -j4
./kitti(or ./sphere)
sudo make install
After compilation, execute
g2oToKitti -n [RobotNumber] -is [inputFilenameSuffix(like _optimized)] -os [outputFilenameSuffix(like _optimized)]
evo_traj kitti [outputFilename(like 0.txt)] -p --plot_mode=xz
For example, if there are files named 0.g2o
, 1.g2o
,0_optimized.g2o
,1_optimized.g2o
, you can execute
g2oToKitti -n 2
g2oToKitti -n 2 -is _optimized -os _optimized
Then you will get 0.txt
,1.txt
,0_optimized.txt
,1_optimized.txt
and execute
evo_traj kitti [yourFiles] -p --plot_mode=xz
x.g2o :initial poses
x_optimized.g2o :optimized poses
x_optimizedTUM.txt :optimized posed(TUM format)
initial.g2o :total initial poses
fullGraph_optimized :optimized and merged poses
centralizedxxx.g2o :merged and optimized poses
You can use distributed_mapper
to optimize them and get optimized result, then use g2oToKitti
for visualization.