RoboND-SLAM-Project

This report studies Simultaneous Localization and Mapping (SLAM) problem using Real-Time Appearance-Based Mapping (RTAB-Map) in a Gazebo environment. A robot model from the previous localization project was used with an added RGB-D camera to map both predefined and custom made environment. With data from robot sensors and odometry data , using RTAB-Map ROS library database of environments were built, and both 2D and 3D maps were generated.Please refer to project writeup for more information.

Project Setup and run

Create catkin workspace, if you haven't already.For this project catkin_ws is used, if your workspace name is different, change the commands accordingly.

$ mkdir -p ~/catkin_ws/src
$ cd ~/catkin_ws/src
$ catkin_init_workspace
$ cd ~/catkin_ws
$ catkin_make

Create package called "slam_project":

$ cd ~/catkin_ws/src
$ catkin_create_pkg slam_project

Copy the content of this repo into slam_project folder, source and build the project:

$ cd ~/catkin_ws
$ source devel/setup.bash
$ catkin_make

To run the mapping process open up 4 terminals and run following commands seperately:

$ cd ~/catkin_ws
$ roslaunch slam_project world.launch
$ roslaunch slam_project teleop.launch
$ roslaunch slam_project mapping.launch
$ roslaunch slam_project rviz.launch

Then on the terminal that run teleop.launch move robot around the map using keyboard to map the world

To change the world environment edit world_name parameter in world.launch file

<include file="$(find gazebo_ros)/launch/empty_world.launch">
  	<!-- <arg name="world_name" value="$(find slam_project)/worlds/kitchen_dining.world"/>-->
    <arg name="world_name" value="$(find slam_project)/worlds/myworld.world"/>
    <arg name="paused" value="$(arg paused)"/>