In robotics, automation, and industrial vision, depth accuracy isn’t optional — it’s mission-critical. But the real world isn’t gentle on cameras. Vibrations, impacts, and heat cycles are unavoidable in the field, and each can gradually push stereo calibration out of alignment. The result? skewed depth maps, unreliable measurements, and systems that no longer see the world as they should.
Calibration drift is simply physics — every stereo system experiences it over time. Traditionally, the only answer has been manual recalibration with tools, targets, and downtime — a disruptive process that often gets delayed until performance has already degraded.
Luxonis Dynamic Calibration changes that. Our on-device calibration health monitoring continuously detects drift and restores factory-level accuracy automatically — in the field, with no setup, no targets, and no downtime. The result is stereo vision that stays reliable where and when it matters most, keeping your systems productive without costly interruptions.
Built for Real-World Scenes
Dynamic Recalibration thrives in the same environments Luxonis devices are built for: factory floors, outdoor robotics, drones, and inspection systems. Everyday textures — from brick walls to cluttered workbenches to tree bark — provide all the features needed to maintain alignment.
No special lighting. No controlled setup. Just depth accuracy that adapts to the real world — delivering higher uptime, lower maintenance effort, and confidence at scale.
TIP: Best results: bookshelves, brick walls, workshop benches, cluttered desks, tree bark, soft daylight. Avoid: blank walls, reflective glass, or repeating identical tiles.
Getting Started: The Easiest Way to Calibrate
The fastest way to get started is by using the latest version of the OAK-Viewer:
👉 Download OAK-Viewer
Once installed:
Switch to the "Dynamic Calibration" profile from the dropdown menu.
The Calibration Quality Check will start automatically (top right corner).
When complete, you'll see whether the camera "Needs Calibration" (upper left corner).
If it does, just click "Recalibrate" (top right) — and you're done.
No setup. No scripts. Just dynamic calibration in a few clicks.
Supported Devices
Dynamic Recalibration is fully supported on:
- ✅ OAK Series 2 full stereo lineup (excluding FFC)
- ✅ OAK Series 4 full stereo lineup.
Any Luxonis stereo camera running DepthAI 3.0 or later can integrate dynamic recalibration immediately.
Developer Quickstart: Plug Dynamic Recalibration Into Your Pipeline
The DepthAI repo includes a minimal working interactive Python example that demonstrates how to run Dynamic Recalibration and integrate it into your own pipeline.
Launch the Example
Clone the repo and run the interactive tool:
git clone https://github.com/luxonis/depthai-core.git
cd depthai-core/
python3 -m venv venv # if you are on Linux
source venv/bin/activate # if you are on Linux
python3 examples/python/install_requirements.py
python3 examples/python/DynamicCalibration/calibration_integration.py
This launches a live demo where you can:
This interactive script provides a clean foundation — ideal for developers who want to explore, test, and extend Dynamic Recalibration in their own embedded or deployed applications.
Please visit our Dynamic Calibration Documentation page to get to know more.
Enhance Depth Accuracy (Optional)
For better performance with active stereo, enable the IR dot projector:
device.setIrLaserDotProjectorIntensity(0.9) # Range: 0 to 1
Let Your Camera Take Care of Itself
With Dynamic Recalibration, your Luxonis device becomes a self-correcting perception system — ideal for real-world conditions where accuracy can’t afford to drift.
No manual resets. No production pauses. Just smarter vision that holds up over time.
Next steps
Here are three easy ways to take the next step with Dynamic Calibration:
📘 Learn More in the Docs
Dive into the Dynamic Calibration documentation to explore usage, configuration, and integration details.
📦 Buy a Compatible Stereo Camera
Head to the Luxonis Shop and pick up an OAK-D or other DepthAI 3.0–supported stereo device to try it hands-on.
🛠️ Build Into Your Product
Use the provided Python example as a foundation and start integrating Dynamic Recalibration directly into your own vision pipeline.
