This fork contains additionnal implementation of the paper: M. Brossard, S. Bonnabel and A. Barrau, Unscented Kalman Filter on Lie Groups for Visual Inertial Odometry, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2018. [HAL paper]
If you use our modified code in your research, please cite the original MSCKF VIO paper and:
@INPROCEEDINGS{2018_Brossard_Unscented,
author = {Martin Brossard and Silvère Bonnabel and Axel Barrau},
booktitle={2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year={2018},
}
The MSCKF_VIO package is a stereo version of MSCKF. The software takes in synchronized stereo images and IMU messages and generates real-time 6DOF pose estimation of the IMU frame.
The software is tested on Ubuntu 16.04 with ROS Kinetic.
Original Video: https://www.youtube.com/watch?v=jxfJFgzmNSw&t=3s Original Paper Draft: https://arxiv.org/abs/1712.00036
Penn Software License. See LICENSE.txt for further details.
Most of the dependencies are standard including Eigen, OpenCV, and Boost. The standard shipment from Ubuntu 16.04 and ROS Kinetic works fine. One special requirement is suitesparse, which can be installed through,
sudo apt-get install libsuitesparse-dev
The software is a standard catkin package. Make sure the package is on ROS_PACKAGE_PATH after cloning the package to your workspace. And the normal procedure for compiling a catkin package should work.
cd your_work_space
catkin_make --pkg msckf_vio --cmake-args -DCMAKE_BUILD_TYPE=Release
An accurate calibration is crucial for successfully running the software. To get the best performance of the software, the stereo cameras and IMU should be hardware synchronized. Note that for the stereo calibration, which includes the camera intrinsics, distortion, and extrinsics between the two cameras, you have to use a calibration software. Manually setting these parameters will not be accurate enough. Kalibr can be used for the stereo calibration and also to get the transformation between the stereo cameras and IMU. The yaml file generated by Kalibr can be directly used in this software. See calibration files in the config folder for details. The two calibration files in the config folder should work directly with the EuRoC and fast flight datasets. The convention of the calibration file is as follows:
camx/T_cam_imu: takes a vector from the IMU frame to the camx frame.
cam1/T_cn_cnm1: takes a vector from the cam0 frame to the cam1 frame.
The filter uses the first 200 IMU messages to initialize the gyro bias, acc bias, and initial orientation. Therefore, the robot is required to start from a stationary state in order to initialize the VIO successfully.
There are launch files prepared for the EuRoC and fast flight dataset separately. Upon launching the msckf_vio_*.launch, two ros nodes are created:
image_processortakes the stereo images to detect and track features.viotakes the feature measurements and tightly fuses them with the IMU messages to estimate pose.
Edit the file msckf_vio_*.launch for selection S-MSCKF, S-UKF-LG, or S-IEKF vio filter.
Subscribed Topics
imu (sensor_msgs/Imu)
IMU messages is used for compensating rotation in feature tracking, and 2-point RANSAC.
cam[x]_image (sensor_msgs/Image)
Synchronized stereo images.
Published Topics
features (msckf_vio/CameraMeasurement)
Records the feature measurements on the current stereo image pair.
tracking_info (msckf_vio/TrackingInfo)
Records the feature tracking status for debugging purpose.
debug_stereo_img (sensor_msgs::Image)
Draw current features on the stereo images for debugging purpose. Note that this debugging image is only generated upon subscription.
Subscribed Topics
imu (sensor_msgs/Imu)
IMU measurements.
features (msckf_vio/CameraMeasurement)
Stereo feature measurements from the image_processor node.
Published Topics
odom (nav_msgs/Odometry)
Odometry of the IMU frame including a proper covariance.
feature_point_cloud (sensor_msgs/PointCloud2)
Shows current features in the map which is used for estimation.