we used limo(https://github.com/johannes-graeter/limo.git) This repository made for visual intelligent term project
src/limo/keyframe_bundle_adjustment_ros_tool/src/mono_lidar/mono_lidar.cpp
src/limo/keyframe_bundle_adjustment_ros_tool/src/mono_lidar/mono_lidar.hpp
src/limo/keyframe_bundle_adjustment_ros_tool/src/common/general_helpers.hpp
sould be changed :
src/limo/keyframe_bundle_adjustment_ros_tool/src/mono_lidar/mono_lidar.cpp
- build :
cd src/limo && bash install_repos.sh
- source update (optional):
cd ~ && source ${your catkin limo dir}/devel_limo_release/setup.bash
- run rosbag :
rosbag play 04.bag -r 0.1 --pause --clock (run roscore)
- run node :
roslaunch demo_keyframe_bundle_adjustment_meta kitti_standalone.launch
-> test.txt file will be created (test.txt : estimated pose data file.)
download (evaluate_odometry.cpp, matrix.cpp, matrix.h, mail.h, 04.txt) in limo directory
required :
pdfcrop : sudo apt-get -y install texlive-extra-utils
gnuplot : sudo apt-get -y install gnuplot
setup :
make resulto4 directory in limo direcotry (cd src/limo && mkdir resulto4)
move test.txt file in the result04 directory
1. g++ -O3 -DNDEBUG -o evaluate_odometry evaluate_odometry.cpp matrix.cpp
2. ./evaluate_odometry 04
-> evaluation files will be created in the result04 directory.