cs4home_person_tracker_module

This repository provides the cs4home_person_tracker ROS 2 package. It maintains person tracks from YOLO detections or from fused camera and laser detections, and publishes the resulting tracks for visualisation and downstream robot tasks.

The existing package directory cs4home_vision_module and dependency manifest thirparty.repos are retained unchanged.

CoreSense role

The terms below follow the CoreSense Ontology (CSO).

• Camera and laser devices are Sensors that acquire observations of people.
• Detection association, filtering and track maintenance form a Cognitive Function.
• The module provides the Cognitive Capability to maintain a person estimate over time.
• That estimate supports Tasks such as person following and human-aware navigation.

Data flow

flowchart LR
    camera["YOLO person detections"] --> tracker["Person tracking cognitive function"]
    laser["Optional laser leg detections"] --> tracker
    tracker --> association["Association and track maintenance"]
    association --> tracks["Tracked-person markers"]
    tracks --> tasks["Following and navigation tasks"]

Loading

The default launch mode uses YOLO detections only. Laser and YOLO fusion can be enabled when the laser people detector and its model are available.

Requirements and build

Use Ubuntu 22.04 and ROS 2 Humble. The workspace needs YOLO ROS, cs4home_architecture, the state observers, the UPO laser people detector for fusion mode, and cs4home_msgs from cs4home_vision_module.

mkdir -p ~/cs4home_ws/src
cd ~/cs4home_ws/src
git clone https://github.com/CoreSenseEU/cs4home_person_tracker_module.git
git clone https://github.com/CoreSenseEU/cs4home_vision_module.git
vcs import --recursive < cs4home_person_tracker_module/thirparty.repos
cd ..
rosdep install --from-paths src --ignore-src -r -y
colcon build --symlink-install --packages-up-to cs4home_person_tracker
source install/setup.bash

Run

Start YOLO and activate its lifecycle nodes:

ros2 launch yolo_bringup yolo.launch.py
ros2 lifecycle set /yolo_node configure
ros2 lifecycle set /yolo_node activate
ros2 lifecycle set /yolo_3d_node configure
ros2 lifecycle set /yolo_3d_node activate

Then launch the tracker:

ros2 launch cs4home_person_tracker person_tracker.launch.py

The tracker launch file configures and activates the person_tracker lifecycle node automatically.