TELLO_SLAM_GAZEBO

ORB_SLAM3 with DJI Tello in Gazebo

🔗 Quick Links

• 🤝 Author
• 📍 Overview
• 👾 Demo
• 🗂️ Repository Structure
• 🚀 Pre-Requisites
• ⚙️ Installation
• ► Running
• 🤖 Extra Instructions
• 📄 License

---

🤝 Author

Yuang Du (ucab190@ucl.ac.uk)

---

📍 Overview

This project is to simulate DJI Tello drone in Gazebo to perform simultaneous localization and mapping (SLAM) using ORB_SLAM3.

---

👾 Demo

Mono SLAM:

Click to watch the video demonstration.

Ape comparison w/wo scale alignment from evo

---

🗂️ Repository Structure

└── tello_slam_gazebo/
    ├── src
    │   ├── [hector_gazebo](git@github.com:alstondu/hector_quadrotor.git)
    │   ├── [hector_localization](git@github.com:alstondu/hector_localization.git)
    │   ├── [hector_quadrotor](git@github.com:alstondu/hector_quadrotor.git)
    │   ├── [orb_slam3_ros](git@github.com:alstondu/orb_slam3_ros.git)
    │   └── [tello_ros_gazebo](git@github.com:alstondu/tello_ros_gazebo.git)
    ├── media
    ├── LICENSE
    └── README.md

---

🚀 Pre-Requisites

See prerequisites of each submodule.

Submodule	Employed system
Gazebo plugins support	hector_gazebo
TF and localization support	hector_localization
Quadrotor controller support	hector_quadrotor
Visual SLAM	orb_slam3_ros
Tello model and discription	tello_ros_gazebo

To evaluate the trajectory obtained, evo is required.

---

⚙️ Installation

git clone --recursive git@github.com:alstondu/tello_slam_gazebo.git
cd tello_slam_gazebo
catkin build

---

► Running

Tip

_{source the workspace in each terminal before entering the command.}

Monocular SLAM

Launch the Main System:

Launch the system in terminal 1:

roslaunch tello_driver tello.launch

Keyboard Control (To Do: Joystick Control)

Run teleop node in terminal 2:

rosrun keyboard_teleop keyboard_teleop_node.py _repeat_rate:=10.0

Save trajectory

Before driving the drone, in terminal 3, record the predicted trajectory(/orb_slam3_ros/camera_pose) and the ground truth(/ground_truth/state) as rosbag:

rosbag record /orb_slam3_ros/camera_pose /ground_truth/state

EVO Evaluation

Convert the saved rosbag to tum format:

evo_traj bag [bag_name] /orb_slam3_ros/camera_pose /ground_truth/state --save_as_tum

Change the suffix of the files from .tum to .txt, plot the trajectories with(-a for alignment, -as for alignment and scale):

evo_traj tum orb_slam3_ros_camera_pose.txt --ref ground_truth_state.txt -a -p --plot_mode=xyz

For APE (Absolute Pose Error), run:

evo_ape tum ground_truth_state.txt orb_slam3_ros_camera_pose.txt -vas -r full -p --plot_mode=xyz

---

🤖 Extra Instructions

Camera Calibration

Launch the world with the calibration board and the tello in terminal 1:

roslaunch tello_driver cam_cal.launch

Run the keyboard teleop node in terminal 2:

roslaunch tello_driver keyboard_teleop_node.launch

Run cam_calibration node:

rosrun camera_calibration cameracalibrator.py --size 7x7 --square 0.25 image:=/front_cam/camera/image camera:=/front_cam

Drive the drone around the board until X, Y, Size, Skew all turn green. Click on the 'CALIBRATE' button, 'Save' the parameters and exit with 'COMMIT'.

---

📄 License

LICENSE: MIT. See LICENSE.txt

DISCLAIMER:

THIS INFORMATION AND/OR SOFTWARE IS PROVIDED BY THE AUTHOR "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS INFORMATION AND/OR SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.