This repository contains ros2 packages.

Object Tracking in ROS2

This project implements real-time object detection and tracking using ROS2 (Humble). The system captures images from a live camera or video, processes them through a YOLO V8 model, and tracks objects using a Centroid Tracking Algorithm.

---

Nodes & Topics

This project consists of the following ROS2 nodes:

Image Publisher (image_publisher)

• Publishes: /camera/image_raw
• Streams images from a live camera or video file.

Image Subscriber (image_subscriber)

• Subscribes: /camera/image_raw
• Publishes: /object_detection
• Runs YOLOv8 to detect objects and outputs bounding boxes.

Object Tracking (object_tracking)

• Subscribes: /object_detection
• Publishes: /object_tracking
• Tracks objects across frames using Centroid Tracking.

Visualization (visualization)

• Subscribes: /camera/image_raw, /object_detection, /object_tracking
• Publishes: /visualization/image
• Overlays detections and tracking results on images.

---

Project Structure

Real-Time-Object-Tracking-ROS2/src/
│── object_tracking_viz/
│   ├── config/                         # YAML configuration files
│   │   ├── params.yaml                 # ROS2 parameters (camera, video file, tracking settings)
│   ├── include/object_tracking_viz/    # Cpp header files
│   │   ├── header.cpp         
│   ├── launch/                         # ROS2 launch files
│   │   ├── app.launch.py               # Launch all nodes
│   ├── media/                          # Sample video files for testing
│   │   ├── test1.mp4
│   │   ├── test2.mp4
│   │   ├── test3.mp4
│   ├── object_tracking_viz/            # Python packages
│   │   ├── __init__.py
│   ├── scripts/                        # Python scripts
│   │   ├── image_subscriber.py         # Subscribes images and detect objects using YOLOv8
│   │   ├── object_visualization.py     # Overlays tracking & detection results
│   ├── src/                            # C++ source files
│   │   ├── image_publisher.cpp         # Publishes images from camera or video
│   │   ├── object_tracking.cpp         # Tracks objects using Centroid Tracking
│   ├── CMakeLists.txt                  # ROS2 build configuration
│   ├── package.xml                     # Package dependencies
│── .gitignore
│── README.md                           # Project documentation

---

Installation

Clone this repository into your ROS2 workspace.

Install the ROS2 Hunble distribution:

https://docs.ros.org/en/humble/Installation.html

Dependencies

This project requires ROS2 Humble, along with additional system and Python dependencies.

System Dependencies

Install the required system packages:

sudo apt update && sudo apt install -y \
    python3-opencv \
    ros-humble-cv-bridge \
    ros-humble-image-transport \
    ros-humble-rclcpp \
    ros-humble-std-msgs \
    ros-humble-ament-index-cpp \
    libeigen3-dev

Python Dependencies

Install the required Python packages:

pip install opencv-python numpy ultralytics nlohmann-json

Build the Package

cd ~/Real-Time-Object-Tracking-ROS2
colcon build --packages-select object_tracking_viz
source install/setup.bash

Configuration (params.yaml)

Modify config/params.yaml to adjust the camera usage, video file name, and tracking settings

Running the Nodes

Use the launch file to start all ROS2 nodes:

ros2 launch object_tracking_viz app.launch.py

OR, manually run each node:

ros2 run object_tracking_viz image_publisher
ros2 run object_tracking_viz image_subscriber
ros2 run object_tracking_viz object_tracking
ros2 run object_tracking_viz object_visualization

---

RQT Graph

Below is the rqt_graph showing the ROS2 nodes and their connections.

---

Demo

---

Resources & References

Here are some helpful resources related to using WSL2, USB support in WSL2, Gazebo Simulation, and OpenCV VideoIO:

ROS2:

• ROS2 Documentation
• Understand ROS2
• ROS2 Resources
• ROS2 Tutorials
• Camera usage in ROS2

WSL2 & USB Camera in WSL2:

• Microsoft Docs - Connect USB in WSL
• WSL2 Linux Kernel (GitHub)
• VideoIO v4l2 /dev/video0 select timeout issue (OpenCV Forum)

Additional Tutorials:

• Package setup for both Cpp and Python
• Launch file for multiple nodes
• YOLOv8 on ROS2
• Multi-threaded execution
• Yaml parameters
• Custom msgs

Gazebo Simulation:

• Gazebo Fortress - Getting Started
• TurtleBot3 Simulation Guide

---

Contributing

Contributions are welcome! Please fork the repository and create a pull request.

👩‍💻 Author

Developed by Shivani Khandelwal.

Connect with me here - Profile

For questions or improvements, feel free to reach out at shivanikhandelwal487@gmail.com.

---