Staff filtering and unique visitors with anonymous re-identification.
A unique ID per person - across cameras, across visits, completely anonymous.
KSI has developed proprietary anonymous re-identification technology that assigns a unique ID to each person, without storing images or personal data. It solves two critical problems at once: automatically filtering out staff so they don't contaminate your metrics, and counting true unique visitors instead of repeat visits from the same customer.
Request a demo →The two problems it solves
Without re-identification, your numbers are lying to you.
Standard people counters cannot tell who is who. They count entries - every entry. That means staff walking in and out look like customers, and a single customer who returns three times in a week looks like three different visitors. Both errors compound silently and break your conversion math.
Staff contaminate your metrics
Without re-identification, every staff member entering and leaving counts as visitor traffic. Your conversion drops artificially and your peak-hour readings are inflated. Filtering by uniform or schedule is unreliable.
Repeat visits inflate your traffic
A loyal customer who visits 3 times this week looks like 3 visitors in your dashboard. You cannot measure real reach, real loyalty, or real conversion per person.
How it works
One person, one anonymous ID - even across cameras and visits.
KSI extracts a morphological signature from each person - anonymous body and gait features, never facial data. The same person captured by two different cameras, or returning some time later, gets matched to the same anonymous ID. The original images are never stored: only the signature and the ID.
What this unlocks
Capabilities only re-identification makes possible.
True unique visits
Stop counting the same person multiple times. Get accurate reach by store, by zone, by mall - and measure real loyalty.
Automatic staff exclusion
No uniforms, no manual tagging. Staff are recognized from their behavior pattern and excluded from conversion math.
Real per-visitor dwell time
Measure how long each individual stayed - not an average smudged across all entries. Spot bouncers, browsers and converters separately.
Full journey reconstruction
If a person passed through this zone, then that one, then left - we know. Reconstruct the actual path each customer took.
Customer-level conversion
Match transactions to behavior at the person level: who looked, who tried, who bought, who came back.
Privacy-safe by design
No facial recognition. No personal data. No image storage. The same anonymous ID is the only thing that travels between cameras.
Side by side
What you measure with vs without re-identification.
| Without re-ID | With KSI re-ID | |
|---|---|---|
| Visitor count | Inflated by staff and repeat visits | True unique visitors, staff excluded |
| Conversion rate | Diluted by non-customer traffic | Real conversion per unique customer |
| Dwell time | Averaged at the door | Measured per individual visitor |
| Customer journey | Aggregate flow only | Reconstructed path per person |
| Privacy compliance | Often relies on facial recognition | Anonymous body signature, GDPR-compliant |
By industry
What re-identification unlocks across industries.
Retail
Real conversion per customer. Bouncing-customer detection. Dwell time per individual. Loyalty measurement based on actual return visits.
Shopping malls
Unique mall visitors vs repeat visits. Path reconstruction across zones, floors and tenants. Per-tenant capture rate from real foot traffic.
Airports
Passenger journey from entry to gate. Real wait-time per passenger across queues. Re-encounter rate at non-aero retail.
Supermarkets
Filter out staff and stockers. Measure shopper dwell vs employee movement. Per-customer basket projection from behavior.
Privacy by design
Anonymous from the first frame to the last byte.
KSI re-identification uses anonymous morphological features only - never facial data. Original images are never stored. The unique ID generated for each person is locally scoped, cannot be reverse-engineered into a personal identity, and complies with GDPR, ISO 27001 and the strictest privacy regulations worldwide.
Anonymous re-identification FAQ
It is a proprietary technology that assigns a unique, anonymous ID to each person based on their morphological signature (body characteristics, gait, clothing, but never facial data). If the same person is captured by two different cameras, or visits the space again later, KSI associates them with the same ID without needing to recognize or identify them.
Yes. It does not use facial recognition, does not store images or personal data, and complies with GDPR, LGPD, CCPA and ISO 27001.
No. The generated ID has local scope and cannot be reversed into a personal identity; it only makes it possible to recognize that it is the same person across different cameras or visits, without knowing who they are.
It makes it possible to automatically exclude staff from the count, measure real unique visitors instead of counting the same person several times, calculate real dwell time per visitor and reconstruct the full journey they took within the space.