Sensor Fusion using Extended Kalman Filter in ROS

This project employs sensor fusion using the EKF ROS package to localize the robot inside a Gazebo environment.

Dependencies

ROS Distro : Noetic
OS : Ubuntu 20.04 LTS

Perform a System Update/Upgrade

$ apt-get update
$ apt-get upgrade -y

Now, Clone this repository in /src of your ROS workspace

$ git clone https://github.com/Apatil10/SensorFusion-EKF.git

Edit the main.launch file

Under /main/launch/ , edit the main.launch file:

<node pkg="rviz" type="rviz" name="rviz" args="-d {update your absolute path for the EKF.rviz config file}"/>

Example: 
<node pkg="rviz" type="rviz" name="rviz" args="-d /home/aditya(user_name))/EKF_project(worksapce)/src/EKF.rviz"/>

Install Packages Dependancies

$ cd {your Catkin_workspace}
$ source devel/setup.bash
$ rosdep install --from-paths src --ignore-src --rosdistro noetic -y

Build the Packages

$ catkin_make
$ source devel/setup.bash

Launch the main file

$ roslaunch main main.launch

Now, you should see Gazebo and rviz launching. Please note that Gazebo might take up to 5 min to launch!

End Result

In the terminal, use the keyboard commands(w-a-s-d-x-space.) and drive the robot around. The red trajectory represents the Odom path whereas the green trajectory represents the EKF path.