AnonymousPersonID
Attribute-based person search in recorded CCTV footage.
- Search recorded footage by clothing colour
- Search by bags, luggage and visible belongings
- A snapshot and record for each person found
CCTV analytics, person search & multi-camera tracking
Site plan · one route
Cam 03Cam 01Cam 02Cam 04
Held byCam 03Cam 01No viewCam 02
One stretch is unwatched and it is drawn, not hidden. The route on either side of it is still the same route.
Overview
What Video Intelligence does
Video Intelligence turns existing CCTV into something that answers questions. It searches recorded footage for a person by what they were wearing and carrying, watches the areas that matter, keeps track of how crowded each one is, and raises an alert that names the place it came from.
Most organisations already have more cameras than anyone can watch. The footage is recorded faithfully and then used only after something has gone wrong, when somebody sits down and scrubs through hours of it looking for a moment they half remember.
Video Intelligence changes what those cameras are for. Areas that matter can be marked and given their own rules, so entering a restricted zone or a crowd building beyond what a space can hold raises an alert while it is still happening. Recorded footage can be searched for a person by what they were wearing and what they were carrying, which turns an afternoon of scrubbing into a question you can ask.
Every alert names the area it came from, which is the difference between useful and ignored. A guard is pointed at one place rather than at a wall of screens, and a limit set for a busy entrance does not have to be the limit for a quiet corridor.
- Products across CCTV monitoring and person search
5
Products across CCTV monitoring and person search
- Every monitored area carries its own rules and limits
Per area
Every monitored area carries its own rules and limits
- An alert identifies the area it came from
Named alerts
An alert identifies the area it came from
In this category
5 products. The same cameras, answering different questions
One product searches footage after the event. The others watch as it happens, for entry into an area that should be empty, for a crowd growing past what a space can take, and for a weapon among people. There is a product for each question.
CrossCam PersonTrack
Live face recognition across your camera network, matched against your watch-lists.
- Live 1:N matching against enrolled watch-lists
- Graded alerts an operator adjudicates, not accepts
- Runs on your own hardware, with no external call
SecureZone Monitor
CCTV-based restricted area and crowd monitoring.
- Mark specific areas as secure zones
- Alert on unauthorised entry to a zone
- Set a crowd limit for each monitored area
CrowdSentinel AI
Multi-camera crowd monitoring and alert system.
- Monitor crowds across multiple CCTV cameras
- Set a separate crowd limit for each area
- Alert when an area goes above its limit
WeaponGuard AI
Weapon detection and crowd safety for retail environments.
- Look for visible weapons in crowded areas
- Continuous crowd monitoring across cameras
- Alert when a potential weapon is identified
How it works
Four steps, and no new cameras
The cameras stay where they are. What changes is that somebody has said which areas matter and what should happen when a rule is broken.
- Live camerasWatched as it happens
- Recorded footageUploaded and searched after
Your existing estate
What you defined
Areas, rules and limits
- Areas marked on the views you already have
- Each area carrying its own rules
- Its own crowd limit, set to what it can hold
- Judged independently of every other area
Every alert names its area
An alert points a guard at one place rather than at every screen, and a limit set for a busy entrance is not the limit for a quiet corridor.
- An alert while it mattersRaised as the situation develops
- Footage you can questionSearched by appearance, not by scrubbing
- 01
Connect what you have
The cameras covering the places that need watching are connected. There is no requirement to replace an estate that already works, and recorded footage can be uploaded for searching after the event.
- 02
Mark what matters
Areas are marked on the camera views you already have. One view can carry more than one area, because a single camera often covers a doorway that must stay clear and a corridor that simply gets busy.
- 03
Set the rules for each one
Each area is given its own rules and its own crowd limit, set to what that space can actually hold. Areas are then judged independently, so a busy entrance does not put a quiet corridor into alarm.
- 04
Get an alert, or ask a question
When a rule is broken an alert is raised while the situation is developing, naming the area it came from. Afterwards, recorded footage can be searched for a person by what they were wearing and carrying rather than watched end to end.
Use cases
Where organisations put it to work
Wherever there is more footage than there are people to watch it. These are the situations teams bring to the category most often.
Airports and transport hubs
Watch crowd levels across entrances, waiting areas and boarding zones, each with the limit that space can take, and get an alert naming the area while numbers are still rising rather than once it is full.
Retail and shopping centres
Keep busy areas in view and raise an alert if a weapon becomes visible among the people present, pointing staff at one part of the store instead of at every screen at once.
Law enforcement and investigations
Search recorded footage for a person by clothing colour and the belongings they were carrying, which turns hours of review into a question and does not require knowing who they are.
Stadiums, venues and events
Monitor crowd levels across separate sections and around exits, so organisers can act on the section that is filling rather than on an overall figure that hides it.
Industrial sites and critical facilities
Mark the areas that should stay empty and be told when someone enters one, while it is happening, rather than discovering it in a review the following week.
Public gatherings and places of worship
Help authorities keep track of crowd levels across large gathering areas and identify the specific places where numbers are becoming a concern.
FAQ
What people ask about Video Intelligence
Answered the way a security manager, an operations lead or a privacy officer asks them.
Video Intelligence is FaceOff's suite for getting answers out of CCTV. It searches recorded footage for a person by what they were wearing and carrying, watches areas you have marked and alerts on unauthorised entry, keeps track of how crowded each area is against a limit you set for it, and looks for visible weapons in busy places. Every alert names the area it came from.
No. The suite is built to work with the cameras organisations already have, because a proposal that begins by replacing an estate is a proposal that does not happen. The cameras covering the places you need watched are connected, and the areas that matter are marked on the views those cameras already produce.
No, and that is the point. The selected areas are watched continuously and an alert is raised when a rule you set is broken, so staff are told about a situation while it is developing instead of counting people on a monitor. Because each alert names its area, attention goes to one place rather than across a wall of screens.
Yes. Recorded footage can be searched by appearance, using clothing colour and the bags or belongings a person was carrying, which means a search can be run when nobody knows the person's name. Each person found is kept as a record with a snapshot and those visual details, so a reviewer can work through candidates rather than through raw footage.
Because spaces are not alike. A crowd that is unremarkable in a concourse is a problem in a stairwell, so a single limit across a site either alarms constantly or misses what matters. Each marked area carries the limit that space can actually hold, and an alert in one area does not depend on how busy another one is.
Video Intelligence is used by airports and transport hubs monitoring passenger crowds, retailers and shopping centres watching busy areas, law enforcement searching recorded footage during investigations, stadiums and event organisers managing crowd levels by section, industrial and critical facilities protecting restricted areas, and authorities overseeing large public gatherings.
Yes. An alert is a reason for someone to look, not a finding of fact, and any method of watching a space, including a person doing it, can be mistaken in either direction. Naming the area an alert came from is part of the answer to that: a guard can confirm or dismiss it in moments because they know exactly where to look, which is what keeps alerts worth responding to.
The rest of the line
More in Data Security
See Video Intelligence run on your own cameras
Bring your camera layout, or footage you have had to search by hand. We will walk through what it takes to set up and what the alerts look like.