Skip to main content

ArUco Detection Setup

This guide gets ArUco detection running on an HVS - Raze.

What You Need​

  • A working HVS - Raze device running a premade image.
  • Printed tags.
  • The dictionary used to print them.

This guide assumes 36h11.

Goal​

You are done when:

  • a stream is Live
  • preview shows an annotated output
  • detections appear when tags are in view

Step 1: Register A Stream​

  1. Go to OS > Devices.
  2. Click Register stream.
  3. Select device/backend/mode.

Recommended defaults:

  • Backend: Libcamera (for OV9782)
  • Format: NV12 when available
  • Decoder: nv12-luma when using NV12

Register.

Step 2: Basic Camera Controls​

  1. Open the stream.
  2. Go to Controls.
  3. Enable noise reduction and set it to Fast (if available).

Step 3: Create Pipeline From Template​

  1. Go to OS > Pipelines.
  2. Create a new pipeline.
  3. Choose Templates and select the ArUco template.

Step 4: Attach To Stream​

  1. Go back to the stream.
  2. Open Pipelines tab.
  3. Add the pipeline.
  4. Drag it into 1x1.
  5. Select output frame for preview.

Step 5: Set Dictionary​

  1. Open pipeline Tune.
  2. Set dictionary to 36h11.

Verify​

  • you see overlays and detections
  • Outputs tab shows non-image outputs when enabled

If you get no detections:

  • fix focus and motion blur first
  • confirm dictionary matches
  • check Systems logs