Drone_sim on Gazebo

---

Introuction

Gazebo simulation of Parrot drone performing human continous tracking using Openpose powered by ROS

Skills Summary:

• ROS
• Gazebo simulator
• Machine Learning
• Computer Vision (Perception)
• OOP programming
• Innovative Prototyping

---

Env

1. ./env.sh
2. rosrun pysdf sdf2urdf.py sjtu_drone.sdf sjtu_drone.urdf

---

Run

1. ./run_gazebo.sh
2. ./run_script.sh

---

Results

1. NVIDIA TRT Pose estimation

2. OpenCV Openpose

---

Reference

1. https://github.com/NVIDIA-AI-IOT/trt_pose
2. https://github.com/tahsinkose/sjtu-drone
3. https://github.com/tahsinkose/openai-ros-gazebo-workout

---

Appendix

---

Sprint Presentation

• Gazebo & ROS Openpose
• ROS Gazebo Quadrotor

---

Commits

Commit Abbreviation	Description
INIT:	Initialised a file
MOD:	Modified a function or something.
ADD:	Added a function or an aspect of something.
FIX:	A quick little fix.
DEL:	Deleted something.
[NOT TESTED]	Flag for if it hasn't been tested.