Zivid

SUMMARY

The Zivid class provides a Python interface for Zivid 3D cameras, supporting single-frame capture of RGB images, depth images, and 3D point clouds, as well as YAML-based settings management and per-acquisition parameter control. It wraps the Zivid SDK to handle device connection, capture configuration, and safe teardown.

Requirements: The Zivid SDK must be installed and a supported Zivid camera must be connected via Ethernet or USB.

Tested devices:

• Zivid 2 M70

Useful Links

• Zivid Knowledge Base
• Zivid SDK GitHub

Installation

Install Medulla first — see Installation.

Then install the Zivid-specific dependencies:

Download and Install Zivid SDK

1. Download and install the latest Zivid SDK from the Zivid Downloads page.
2. Install Zivid Studio, which is included with the SDK.
3. Connect your Zivid camera.
4. Open Zivid Studio to verify the installation:
   • Confirm the camera is detected and listed.
   • Capture a test frame to validate the connection.
   • Tune 3D and 2D acquisition settings (aperture, exposure, gain, brightness).
   • Extra: Export your settings as a .yml file via File → Export Capture Settings for use with load_settings.

Zivid Studio after successfully installing and connecting the camera.

Python Packages

Install the Zivid dependencies via the Telekinesis SDK:

pip install telekinesis-ai[medulla-zivid]

Getting Started

WARNING

Examples were tested on Windows 11 only.

After verifying the installation with Zivid Studio, make sure to close the studio or disconnect the camera within it before running any medulla script.

WARNING

Only one program can access the camera at a time. Make sure to disconnect from active cameras in Zivid Studio before running the scripts.

Capture a point cloud with a Zivid camera from medulla:

from telekinesis.medulla.cameras import zivid

camera = zivid.Zivid(name="my_zivid_camera")
camera.connect()
pc = camera.capture_pointcloud()
print(pc.positions)
camera.disconnect()

Settings Management

Zivid cameras are configured through settings files exported from Zivid Studio. You can load and save these .yml files at runtime:

# Load settings exported from Zivid Studio
camera.load_settings("settings_inspection_far.yml")

# Fine-tune individual parameters
camera.set_parameter("aperture", 5.6)
camera.set_parameter("color_gain", 2.0)

# Save the modified settings for later use
camera.save_settings("my_custom_settings.yml")

The following parameters can be read and written via get_parameter / set_parameter:

Parameter	Type	Scope
aperture	float | int	Per 3D acquisition
exposure_time	datetime.timedelta	Per 3D acquisition
gain	float | int	Per 3D acquisition
brightness	float | int	Per 3D acquisition
color_aperture	float | int	Per 2D acquisition
color_exposure_time	datetime.timedelta	Per 2D acquisition
color_gain	float | int	Per 2D acquisition
color_brightness	float | int	Per 2D acquisition
noise_threshold	float | int	Global (all captures)
outlier_threshold	float | int	Global (all captures)

For HDR captures with multiple acquisitions, use the acquisition_index argument to target a specific acquisition.

Examples

Skill	Description
Connect	Open a connection to the camera hardware
Disconnect	Release the camera connection and free resources
Capture Color Image	Capture one colour image and visualise with Rerun
Capture Depth Image	Capture one depth frame with colour map and visualise with Rerun
Capture Point Cloud	Capture one colour-mapped 3D point cloud and visualise with Rerun
Load Settings	Load a settings YAML file exported from Zivid Studio
Save Settings	Save the current settings to a YAML file
Get Parameter / Set Parameter	Read or write acquisition parameters such as aperture and gain
Get Intrinsics	Retrieve camera intrinsic parameters for calibration workflows