containerized person recognition with openpose in ROS.
- Docker >= 20.10
- nvidia-container-toolkit
- NVIDIA GPU
- ≈ 20GB of free space
This image can be pulled from DockerHub but it also can be build locally.
This step can be skipped if the image is downloaded.
Navigate to the cloned repository and run
docker build -t ros-openpose-docker .
after this the image is build and saved on the drive.
xhost +local:root; docker run -ti --rm --gpus all --net=host -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix:rw ros-openpose-docker
With each start of the container, the hostname & IP of the ROS slave have to be added to the ects/hosts file with nano. Both IP addresses should be static.
In my case this is:
10.0.106.30 rosberry
After this we can start the roscore.
To start Openpose or other services in the container we need to open a new terminal. To get into the running Docker Container use:
docker exec -it <YOUR-BUILT-CONTAINER-NAME> bash
Navigate to /openpose/build/examples/tutorial_api_python and start ros_openpose.py