zed-ros2-wrapper
ROS 2 wrapper for Stereolabs ZED cameras, providing access to depth, point clouds, object detection, and spatial mapping.
What is zed-ros2-wrapper?
ZED ROS 2 Wrapper is an open-source bridge that allows Stereolabs ZED stereo cameras to be used within the ROS 2 (Robot Operating System 2) environment. It provides ROS 2 nodes that publish camera data—such as images, depth maps, 3D point clouds, and object detections—as standard ROS 2 topics and services, enabling robotics developers to integrate advanced stereo vision capabilities into their autonomous systems.
Robotics engineers and researchers developing ROS 2-based applications that require stereo vision, depth perception, 3D mapping, or object detection, particularly those using Stereolabs ZED cameras.
Developers choose this wrapper because it provides a well-maintained, feature-rich interface between the powerful ZED SDK and the ROS 2 ecosystem, saving significant development time and ensuring reliable, high-performance data streaming for perception tasks.
Overview
ROS 2 wrapper for the ZED SDK
Use Cases
Best For
- Building autonomous robots that require real-time depth perception and 3D mapping
- Developing ROS 2 applications with object detection and body tracking using ZED cameras
- Integrating stereo vision with GNSS for precise outdoor localization
- Testing perception algorithms in simulation with NVIDIA Isaac Sim
- Creating multi-camera robotic systems with synchronized stereo vision
- Implementing custom object detection models in a ROS 2 pipeline
Not Ideal For
- Projects using ZED X One or first-generation ZED cameras that require advanced features like object detection or spatial mapping
- Teams not committed to Ubuntu 20.04/22.04/24.04 with specific ROS 2 distributions (Foxy, Humble, Jazzy)
- Applications needing a lightweight, dependency-free vision solution without proprietary SDKs or CUDA
- Environments where real-time performance is critical without NVIDIA GPU acceleration
Pros & Cons
Pros
Compatible with a wide range of ZED cameras from ZED to ZED X One, as specified in launch parameters, enabling flexibility across hardware generations.
Publishes color/depth images, 3D point clouds, odometry, and object detections as standard ROS 2 topics, facilitating rich perception pipelines without extra coding.
Includes native modules for object detection, body tracking, and spatial mapping, reducing the need for external packages and simplifying development.
Supports NVIDIA Omniverse Isaac Sim for testing and integrates with NVIDIA Isaac ROS for accelerated data streaming, enhancing deployment options.
Cons
Relies on the closed-source ZED SDK and CUDA, which adds installation complexity, limits portability, and ties users to vendor-specific updates.
Advanced features like object detection and simulation mode are unavailable for ZED X One and older cameras, restricting functionality based on hardware.
Requires building from source, managing ROS dependencies, and tuning DDS settings, as noted in the installation guide, which can be error-prone for newcomers.
Optional features like Point Cloud Transport are not installed by default, requiring additional steps and potentially breaking workflows if overlooked.
Frequently Asked Questions
Related Projects
Universal RobotsUniversal Robots ROS2 driver supporting CB3 and e-Series
odrive_ros2_controlODrive driver for ros2_control
ros2_intel_realsenseThis project is deprecated and no more maintained. Please visit https://github.com/IntelRealSense/realsense-ros for ROS2 wrapper.
ros2_ouster_driversROS2 Drivers for the Ouster OS-0, OS-1, and OS-2 Lidars
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.