> CAPABILITY / 02_REID

Person Re-Identification.

Tracks individuals across cameras without relying on facial biometrics.

Identity & Access

REQUEST_DEMO ALL_CAPABILITIES

MODULE 02_REID

READY

Track ID: · · ·
Cameras visited: · · ·
First/last seen: · · ·
Time per zone: · · ·

EDGE-NODE-01 · ON-PREM5 DATA FIELDS

> THE_PROBLEM

Why it matters.

Faces are not always visible: angled cameras, helmets, masks. Following one person across a site by scrubbing footage takes hours.

> HOW_IT_WORKS.exec()

How it works.

DETECT

Each person and build an appearance signature (clothing, build, colors).

MATCH

Signatures across cameras on the edge box.

TRACE

The route as a timeline of cameras and zones.

LIVE

RUNNING

T+000 [DETECT] CAM_04 · each person and build an appearance signature (clothing, build, colors)

> OUTPUT.schema()

What you get.

Data it produces

Track IDCameras visitedFirst/last seenTime per zonePath timeline

> SCENARIOS.list()

Where it helps.

Follow a subject of an incident across the site

Visitor flow analysis

Find coverage gaps between cameras

Used inGovernment & Smart City

> REQUIREMENTS.check()

Camera & deployment.

Full-body view
Adjacent or overlapping coverage, ideally within ~10–20 m of each other

Privacy & compliance

Processing runs on-premise, inside your network, and video does not leave it. Access is controlled with role-based permissions and audit logs. Use should follow your internal policy and the data protection rules that apply to you, such as Indonesia's PDP Law.

Made to fit your use case.

We train custom models for your exact use case. For example: tune the signature for uniforms, so people in identical work clothes can still be told apart.

CUSTOM_AI

> RELATED.list()