Spatial Perception

Spatial perception combines depth, motion tracking, and on-device AI to understand where things are in the real world—not just in pixel space.

Use cases

Metric 3D awareness for AR overlays, robotic navigation, inspection, and digital twins
Robust pose estimates in dynamic scenes by blending visual features and IMU
Fully on-device: low latency, privacy-friendly, resilient to poor connectivity

How it works

The main way to implement spacial perception in DepthAI is by implementing the SpatialDetectionNetwork which is essentially a combination of the DetectionNetwork and SpatialLocationCalculator.

Code-less quick start with App Store

Spatial Detections App

You can install the Spatial Detection App through our App Store. To install it you don't need to write a single line of code. This is the quickest way to try spatial detection in action.

Click here to see it in Hub

Build your own

Guides and examples

Spatial Detection Network example

Detect objects with YOLO neural network and determine their real locations.

Collision avoidance example

Alert user of collision with camera.

Spatial detection fusion example

Detect object with multiple cameras.

Need assistance?

Head over to Discussion Forum for technical support or any other questions you might have.

ON THIS PAGE