limhaeryong / eskf-gnss-imu-localization

GNSS-IMU Localization based on Error-State Kalman Filter

ESKF BASED GNSS-IMU LOCALIZATION

This repository contains GNSS-IMU Localization based on Error-State Kalman Filter.

The implementation of ESKF is based on Quaternion kinematics for the error-state Kalman filter.

OpenGL is used to visualize GNSS/Filtered Position in realtime and 3D space.

Installation

Dependencies

• Ubuntu 20.04
• ROS2(foxy)
• Eigen3
• GeographicLib
• yaml-cpp
• OpenGL
• nmea_msgs(ROS package)
• nmea_navsat_driver(ROS package)

Install EU Long-term Dataset with Multiple Sensors for Autonomous Driving

EU Long-term Dataset with Multiple Sensors for Autonomous Driving

• Download 2019-01-31 (Fri, snow) None Image rosbag

Convert rosbag1 to rosbag2

• rosbag convert

rosbags-convert utbm_robocar_dataset_20190131_noimage.bag

Clone the repository

cd /your/workspace/src
git clone https://github.com/LimHaeryong/eskf-gnss-imu-localization.git

Modify rosbag path

• modify launch/eskf_gnss_imu_localization.launch.py

    play_rosbag = ExecuteProcess(
        cmd=['ros2', 'bag', 'play', 'change/to/your/rosbag/path']
    )

Colcon Build

cd /your/workspace
colcon build
source ./install/local_setup.bash

How to run

ros2 launch eskf_gnss_imu_localization eskf_gnss_imu_localization.launch.py 

Result

• GNSS position(Red)
• Filtered position(Green)

eskf-gnss-imu-localization's Issues

LC & TC

Hello Lim,

I hope you are doing well. This repository is very beneficial for navigation. I have a question: Did you opt for a loosely or tightly coupled approach when implementing this algorithm?

About IMU Calibration

Hello, thanks to the good package, I've taken a good look at it, but I just have a few questions, so I'm posting it like this

This package contains functions that are predicted by IMU and updated to GNSS. In this process, I wonder if only GNSS is corrected with IMU data? Or if IMU data is also corrected. The reason for asking is that both my IMU and GPS have a lot of errors, so I wonder if they correct the accumulated error of IMU well.