We did the client setup on a NanoPi NEO Plus 2 which was operating Ubuntu Core 20.04 LTS. You can set up a NanoPi by following this tutorial: NanoPi NEO Plus2. After Ubuntu installation, assign an IP to the NanoPi and access it using ssh, then follow these commands:
-
Make sure that you are in root directory.
-
Run:
mkdir opencv_com -
Run:
cd opencv_com -
Run:
sudo nano requirements.txtCopy the text from requirements.txt to this file and save. -
Run:
pip install -r requirements.txt(All libraries will be automatically installed) -
Run:
sudo nano obs_pi.pyCopy content of obs_pi.py to this file. Save and close. (C+S then C+X)
Error: ImportError: libGL.so.1: cannot open shared object file: No such file or directory
- Run:
sudo apt-get update && apt-get install -y python3-opencv
This code will start moving the SIMBot while avoiding obstacles and publishing camera data on a ROS topic which the server will subscribe to
- Make sure you are in opencv_com directory.
- Run:
python3 obs_pi.py
To setup the server, we need to install CCM-SLAM on the PC. After installation:
Make multiple copies of the Client.launch file according to the number of robots being used.
2. Run roslaunch ccmslam Client.launch
3. Run rviz for visualization of the trajectories of multiple agents.
The final setup should look like this:
