Open3DExplorer

Whole-Body Motion Planning and Tracking of a Mobile Robot with a Gimbal RGB-D Camera for Outdoor 3D Exploration

1. Introduction

This work Open3DExplorer is oriented to the rapid autonomous reconstruction task of mobile robots, and for the problems of the field environment, such as openness, uneven texture distribution and rugged terrain, which seriously affect the efficiency and robustness of the autonomous reconstruction of robots, we use the active gimbal and the multi-layer map information to plan the motion of the robot and the camera, and realize an efficient and robust autonomous reconstruction scheme.

Authors: Zhihao Wang, Haoyao Chen from the Networked RObotics and Sytems Lab, and Mengmeng Fu.

If you use Open3DExplorer for your academic research, please cite the following paper [pdf].

@ARTICLE{wang_whole_2023,
  author={Wang, Zhihao and Chen, Haoyao and Fu, Mengmeng},
  journal={Journal of Field Robotics}, 
  title={Whole-Body Motion Planning and Tracking of a Mobile Robot with a Gimbal RGB-D Camera for Outdoor 3D Exploration}, 
  year={2023},
  volume={},
  number={},
  pages={},
  doi={10.1002/rob.22281}}

1.1. Related Video:

Video links: Youtube or Bilibili.

2. Installation

Tested on Ubuntu 18.04.

• Install ROS packages

sudo apt-get install ros-(rosversion)−octomap−rosros−(rosversion)-octomap-ros ros-(rosversion)-octomap-msgs ros-(rosversion)−yocs−cmd−vel−muxros−(rosversion)-yocs-cmd-vel-mux ros-(rosversion)-pcl-ros

• Create a conda environment {YOUR_ENV_NAME} with python3 (python3.6 in my computer) and install below packages, the version of these packages are adapted to yourself cuda version.

conda install cudatoolkit cudnn pytorch torchvision torchaudio pyyaml opencv-python opencv-python-headless scikit-image protobuf segmentation_models_pytorch albumentations 

• Download code

cd {YOUR_PATH}
git clone https://github.com/HITSZ-NRSL/Open3DExplorer.git
cd Open3DExplorer
catkin build

• Install ORB-SLAM

cd Open3DExplorer/src/semantic_slam/ORB_SLAM2
./build.sh

If there is any problems in installing ORB-SLAM, you can refer to ORB-SLAM2.

3. Run examples

Simulator

roslaunch gazebo_continuous_track_example gazebo_tracked_vehicle.launch

SLAM

roslaunch semantic_slam slam_gazebo.launch

Semantic Mapping

conda activate {YOUR_ENV_NAME}
roslaunch semantic_slam semantic_mapping_py3_gazebo.launch

Planning

roslaunch path_planning path_planning_plane.launch

Tracking

roslaunch mpc_tracker mpc_test.launch

4. Acknowledgments

Semantic SLAM in Open3DExplorer is based on Semantic SLAM. This work was supported in part by National Natural Science Foundation of China and Shenzhen Science and Innovation Committee.

LICENSE

The source code is released under GPLv2 license.