M2CS Player

m2cs_player is a ROS1 + Qt5 offline multi-sensor playback tool. It can publish:

• GNSS/INS (/gwnav/nav, /gwnav/gps, /gwnav/imu)
• Livox left/right point clouds and IMU (/livox_left/*, /livox_right/*)
• Ouster point cloud and IMU (/ouster/*)
• Camera image (/camera/image_raw)
• Simulated clock (/clock)

It also provides a GUI for play/pause/loop/stop-skip/sensor toggles, and uses m2cs_player.launch to start RViz and image_proc debayer nodelet.

---

1. Project Info

• ROS package name: m2cs_player
• Executable: m2cs_player
• Launch file: launch/m2cs_player.launch
• Maintainer: Huanfeng Zhao (zhaohf24@mails.jlu.edu.cn)

---

2. Dataset Download

You can download the dataset via the links below:

• google
• baidu (extract code: m2cs)

---

3. Dependencies

Recommended environment: Ubuntu + ROS1 (catkin workspace).

2.1 ROS dependencies (declared in package.xml)

• roscpp, rospy, std_msgs, geometry_msgs
• image_transport, cv_bridge
• pcl_ros, pcl_conversions, pcl_msgs
• camera_info_manager, tf
• cmake_modules

2.2 Third-party dependencies (used in CMakeLists.txt)

• Eigen
• PCL
• OpenCV
• Qt5 (Widgets, Gui, OpenGL)

2.3 Extra message packages used in source code

The source includes these custom message packages. Make sure they are available in the same workspace:

• gwnav_ros_driver
• livox_ros_driver

---

4. Build

Use the following workflow in your catkin workspace.

1. Create a workspace (skip if you already have one):

mkdir -p ~/m2cs_player_ws/src

2. Clone the repository into src:

cd ~/m2cs_player_ws/src
git clone https://github.com/InRobots/m2cs_player.git

3. Build with catkin:

cd ~/m2cs_player_ws
catkin_make

4. Source the workspace environment:

For zsh:

source ~/m2cs_player_ws/devel/setup.zsh

For bash:

source ~/m2cs_player_ws/devel/setup.bash

---

5. Run

roslaunch m2cs_player m2cs_player.launch

This launch file starts:

1. m2cs_player main application
2. rviz (config: rviz/m2cs.rviz)
3. nodelet + image_proc/debayer (namespace defaults to camera)

---

6. Dataset Directory Layout

After clicking Load in the GUI, the selected dataset root should contain at least:

<dataset_root>/
├── data_stamp.csv
├── ins_data/
│   ├── gwnav_nav.csv
│   ├── gwnav_imu.csv
│   ├── gwnav_gps.csv
│   ├── livox_left_imu.csv
│   ├── livox_right_imu.csv
│   └── ouster_imu.csv
├── calib_data/
│   ├── livox_left_to_ouster.yaml
│   ├── livox_right_to_ouster.yaml
│   ├── camera_to_ouster.yaml
│   ├── ouster_to_gwnav.yaml
│   └── camera.yaml
├── lidar_data/
│   ├── livox_left/*.bin
│   ├── livox_right/*.bin
│   └── ouster/*.bin
└── image_data/*.png

If path or files are invalid, the program prints: Please check file path. Input path is wrong.

---

7. GUI Usage

1. Click Load and select the dataset root folder.
2. Click Play to start playback (click again to end).
3. Click Pause/Resume to pause/resume.
4. Drag the slider to jump in timestamp progress.
5. Use checkboxes for options:
   • Loop: restart playback when finished
   • Skip stop section: skip stationary segments
   • Auto start: auto-play after loading
   • Sensor switches: TF, Nav, Ouster, Left Livox, Right Livox, Camera
6. Set Speed playback rate (0.01 ~ 20.0).

---

8. Troubleshooting

8.1 Build error: missing gwnav_ros_driver or livox_ros_driver

Put the required ROS packages in the same m2cs_player_ws/src and rebuild.

8.2 No image or point cloud during playback

• Verify dataset layout and filenames in Section 6
• Check whether corresponding sensor checkboxes are enabled
• Check RViz topics and fixed frame

---

Acknowledgements

• Complex Urban Dataset File Player

---

Citation

@article{zhao2025m2cs,
   title={M2CS: A Multimodal and Campus-Scapes Dataset for Dynamic SLAM and Moving Object Perception},
   author={Zhao, Huanfeng and Yao, Meibao and Zhao, Yan and Jiang, Yao and Zhang, Hongyan and Xiao, Xueming and Gao, Ke},
   journal={Journal of Field Robotics},
   volume={42},
   number={3},
   pages={787--805},
   year={2025},
   publisher={Wiley Online Library}
}