Scan & Count

The problem

Operators want a running sense of how many packages a sweep has covered.

During a bulk sweep, the same label can appear across many video frames. Without deduplication, the frame count is mistaken for how many packages were received, which inflates inventory and hides missed items.

How it works

Video-frame deduplication. The core suppresses duplicate frames for the same decoded symbol within a scan session. This is frame-level deduplication of captured imagery, not a physical count of packages on a pallet.

Operator steps

1. Sweep

   Run a live camera sweep across the shipment. Frames are processed as they arrive.

2. Deduplicate frames

   Repeated frames showing the same decoded symbol on the same package are collapsed so one package is not counted many times from one angle.

3. Review the distinct-package count

   The surface shows how many distinct labels have been observed, as an aid for the operator to confirm coverage. The operator confirms the receipt.

Operator journey

Outputs

• Distinct observed labels within the sweep
• Suppressed duplicate frames for each label
• Operator-facing distinct-package indicator (receiving aid)

Honest limitations

• Video-frame deduplication is not physical quantity. A single package seen from several angles, or several identical packages each seen once, are not the same as a counted quantity. The distinct-label indicator is a receiving aid, not an inventory quantity.
• Vision Scan does not claim a supported mode that tallies a physical package count from imagery. Quantity of received items is confirmed by the operator and the receiving workflow, not inferred by the core.
• Designed for visible, sufficiently resolved labels.

Add scan and count to your receiving stack.

• Canonical JSON
• Android quickstart
• API overview
• View the full workflow page — the public product page for this workflow.