Stereolabs Zed Camera

Provides quick access and useful resources for Stereolabs Zed Camera.

Quick AccessPermalink

• Stereolabs on GitHub
• Examples on GitHub
• ZED SDK
• Install the ZED Python API
• Code Samples
• Documentation
• Getting Started
• Installation Quick Start Guide
• API Reference
• ZED FAQ
  • What is the camera focal length and field of view?
• ZED Troubleshooting
• StereoLabs Forum
• ZED ROS Wrapper
• ZED Docker Image

Before PurchasingPermalink

Check the specs on the store page, and decide which model to purchase.

In general, ZED 2i > ZED 2 > ZED-M > ZED 1.

Quick StartPermalink

We assume you are using Linux.

1. Follow this Quick Start Guide and download the SDK with preferred CUDA version.
2. Install NVIDIA CUDA, with the same version of ZED SDK, and reboot.

3. Install the ZED SDK, with all Yes. (If you don’t need the AI models, you can skip them.)

   sudo apt install zstd
   chmod +x ZED_SDK_Ubuntu18_cuda11.7_v3.8.0.zstd.run
   ./ZED_SDK_Ubuntu18_cuda11.7_v3.8.0.zstd.run
   

4. Verify the SDK is installed with /usr/local/zed/tools/ZED_Explorer.
5. Install the Python package.
6. Download and run the Camera Control example.

More details are provided in the Getting Started page.

Side Note: If you have trouble updating NVIDIA driver, please refer to this post.

Environment SetupPermalink

CUDA CompatibilityPermalink

• Set up ZED and PyTorch.
• Set up ZED and TensorFlow.

ROSPermalink

• ZED ROS Wrapper
• ZED ROS Examples

ZED in Virtual EnvironmentPermalink

Activate the virtual environment (virtualenv, conda, etc.) and run:

python /usr/local/zed/get_python_api.py

Test with:

python -c 'import pyzed'

TroubleshootingPermalink

If you encountered issue when using ZED Explorer, launch the Diagnostic tool: /usr/local/zed/tools/ZED_Diagnostic.

If you encountered problem during Python package installation, see this page.

If you encountered further issues, see the official troubleshooting page.

If you have questions, ask on the StereoLabs Forum.