Modified version of VINS-Mono (commit 9e657be)
GPL-3.0 License
Modified version of VINS-Mono (commit 9e657be on Jan 9, 2019), a Robust and Versatile Monocular Visual-Inertial State Estimator.
VINS-Mono uses an optimization-based sliding window formulation for providing high-accuracy visual-inertial odometry. It features efficient IMU pre-integration with bias correction, automatic estimator initialization, online extrinsic calibration, failure detection and recovery, loop detection, and global pose graph optimization, map merge, pose graph reuse, online temporal calibration, rolling shutter support.
[1] VINS-Mono: A Robust and Versatile Monocular Visual-Inertial State Estimator [2] Online Temporal Calibration for Monocular Visual-Inertial Systems
tested on Ubuntu 16.04 (ROS Kinetic) and Ubuntu 18.04 (ROS Melodic)
Eigen 3.3.3
catkin_make -j2
# or
catkin build
roslaunch vins_estimator euroc.launch
rosbag play <YOUR_PATH_TO_DATASET>/MH_01_easy.bag
with MYNTEYE-S1030
roslaunch mynt_eye_ros_wrapper mynteye.launch
roslaunch vins_estimator mynteye_s1030_mono.launch
Ubuntu 16.04 下 VINS-Mono 的安装和使用(RealSense ZR300)
roslaunch maplab_realsense maplab_realsense.launch
roslaunch vins_estimator realsense_fisheye.launch
sudo usermod -aG docker $YOUR_USER_NAME
cd docker
make build
./run.sh LAUNCH_FILE_NAME # ./run.sh euroc.launch
./run.sh LAUNCH_FILE_NAME
after your changesEvaluate the output trajectory vins_result_loop.tum with ground truth trajectory in the standard dataset (e.g. for EuRoC MAV dataset, the ground truth file is <sequence>/mav0/state_groundtruth_estimate0/data.csv
) using the evo tools.
copy the ground truth file data.csv to the directory as same to vins_result_loop.tum
evaluate (APE & RPE)
evo_traj euroc data.csv --save_as_tum # --> data.tum
evo_ape tum data.tum vins_result_loop.tum --align --plot
evo_rpe tum data.tum vins_result_loop.tum --align --plot
# or
evo_ape euroc data.csv vins_result_loop.tum --align --plot
evo_rpe euroc data.csv vins_result_loop.tum --align --plot
get results
APE w.r.t. translation part (m)
(with SE(3) Umeyama alignment)
max 0.157368
mean 0.081223
median 0.076672
min 0.021434
rmse 0.086200
sse 7.809322
std 0.028865