CaBot People repo

Package	Description
cabot_people	launch file and script for launching people tracking nodes
track_people_cpp	detect and track people (cpp implementation) for improved performance
track_people_msg	msgs for track_people packages
track_people_py	detect and track people

Test

Preparation

• run the script to download dependencies

./setup-dependency.sh

• assume you have docker (Nvidia docker) and docker compose

• make sure you have a PC with a NVIDIA GPU, or a Jeston (Xavier, Orin, Xavier NX)

• if you pull the latest docker images from docker hub, run the following command

docker compose --profile build pull

• if you build docker image, run the script to build image

./bake-docker.sh -i         # run docker image build for your platform

• if you run in development mode, run the script to build workspaces

./build-workspace.sh        # run workspace build
./build-workspace.sh -d     # run workspace debug build (symlink-install)

Bringup Realsense(s), detection, and tracking

• connect realsense(s) to your PC
• edit .env file to specify Relasense serial numbers (required if you use multiple)

CABOT_REALSENSE_SERIAL_1=
CABOT_REALSENSE_SERIAL_2=
CABOT_REALSENSE_SERIAL_3=

• run one of the following script after the build

docker compose -f docker-compose-test-rs3.yaml --profile prod up rs1-prod track-prod           # 1 Realsense in production mode
docker compose -f docker-compose-test-rs3.yaml --profile prod up rs1-prod rs2-prod track-prod  # 2 Realsenses in production mode
docker compose -f docker-compose-test-rs3.yaml --profile prod up                               # 3 Realsenses in production mode

docker compose -f docker-compose-test-rs3.yaml --profile dev up rs1-dev track-dev           # 1 Realsense in development mode
docker compose -f docker-compose-test-rs3.yaml --profile dev up rs1-dev rs2-dev track-dev   # 2 Realsenses in development mode
docker compose -f docker-compose-test-rs3.yaml --profile dev up                             # 3 Realsenses in development mode

docker compose -f docker-compose-test-rs3-framos.yaml --profile prod up rs1-framos-camera-prod rs1-framos-detection-prod track-framos-prod                                                      # 1 FRAMOS in production mode
docker compose -f docker-compose-test-rs3-framos.yaml --profile prod up rs1-framos-camera-prod rs1-framos-detection-prod rs2-framos-camera-prod rs2-framos-detection-prod track-framos-prod     # 2 FRAMOSes in production mode
docker compose -f docker-compose-test-rs3-framos.yaml --profile prod up                                                                                                                         # 3 FRAMOSes in production mode

docker compose -f docker-compose-test-rs3-framos.yaml --profile dev up rs1-framos-camera-dev rs1-framos-detection-dev track-framos-dev                                                  # 1 FRAMOS in development mode
docker compose -f docker-compose-test-rs3-framos.yaml --profile dev up rs1-framos-camera-dev rs1-framos-detection-dev rs2-framos-camera-dev rs2-framos-detection-dev track-framos-dev   # 2 FRAMOSes in development mode
docker compose -f docker-compose-test-rs3-framos.yaml --profile dev up                                                                                                                  # 3 FRAMOSes in development mode

Check /people topic

• check if /people topic is published and recognize someone in front of the camera
• check if /people topic is published at 10-15 hz per camera, if you have two cameras it publishes at 20-30 hz.

docker exec -it $(docker ps -f name=cabot-people-rs1 -q) bash -c "source install/setup.bash && ros2 topic echo /people"
docker exec -it $(docker ps -f name=cabot-people-rs1 -q) bash -c "source install/setup.bash && ros2 topic hz /people"

License

MIT License