F3: Find, Fix and Follow

Viseum Global F3 is find, fix and follow software that tracks a person or a vehicle across a whole camera network, live or in recorded video. It is for police forces, city control rooms, border agencies, airports and critical sites that need to know where someone has been, where they are now and where they are going. It works with the cameras you already own and with Viseum moving cameras.

“The best video intelligence for public safety” (Police Report). A world first from Viseum, Global F3 plays a key role in the emergency chase of suspects and gives the most accurate situational awareness video intelligence.

Global F3 connects to your command and control system. It uses the types and locations of your cameras, their transmission rates and recording media, the threat level of the incident and the likely pursuit speed. From this it works out where the person could be, where they are most likely to be, and which cameras to search first.

Multi-camera person tracking across any brand of camera

Global F3 is designed to work with analogue and digital cameras from many makers, so a journey can be rebuilt across the whole camera network. It registers people as they enter the camera network and re-identifies them as they move between locations. The result is a journey reconstruction: each sighting, with its time, place and video, shown in order.

For investigations, this replaces days of searching stored video for a few seconds of evidence. For live incidents, it points the operator to the next likely camera before the person reaches it.

During investigations, millions of dollars are commonly wasted each year searching stored video. Even with a low-cost video surveillance system, Viseum iVOS Global F3 still delivers the best results.

Person re-identification with a ranked gallery

People in a crowd often look alike. When the person of interest is among similar-looking people, Global F3 shows the operator a gallery of the most likely matches, ranked from most to least likely. The operator examines them and confirms the right one. The software narrows the search; a trained person makes the decision.

Live pursuit with the Viseum moving camera

A fixed camera records what passes in front of it. Viseum adds a moving camera. On the Viseum 360 Security Camera, wide-view cameras detect the person and an automatic moving camera turns, zooms and follows them until a usable image of the face is captured. Viseum has developed this approach since its Tag and Track prototypes of 2005 and 2006.

One Viseum camera does the work of 6 constantly manned cameras, a Metropolitan Police surveillance officer reported.

Alarms go to command and control as audio, text or a visual display, and by SMS or email to staff on the ground.

One picture from many sources

Global F3 can be joined with other Viseum and customer systems:

• Face recognition: matches a face captured during the pursuit against a lawfully held watchlist, for the operator to confirm.
• ANPR and traffic management: number plate reads become sightings in the same journey.
• Public evidence upload: members of the public can send video, images and audio with location data to an investigation.
• Surveillance radar and counter drone alarms: a radar target can be handed to the cameras and followed.
• Mobile phone networks: for authorised agencies that lawfully obtain communications data under their own legal powers (in the United Kingdom, the Investigatory Powers Act 2016), Global F3 provides a surveillance workflow that cross-checks video against the locations and times of callers and recipients.
• Social media investigations: Global F3 can analyse images and text from social media sources the customer is lawfully authorised to access, for each investigation, to corroborate or identify the location and cause of public incidents. It does not build face databases by untargeted scraping of the internet.

Supplied only to authorised agencies, and only where the law of the customer’s country permits it. Searches use lawfully obtained sources under the customer’s legal authority.

The same engine can also trace movements for missing-person searches and major incident reunification, where the law allows it.

Public Safety Video Intelligence

Locating and following suspects during emergencies means studying every public safety AI surveillance camera intensely, yet even the best operators cannot watch every screen for every hour of a shift, and wages are an ongoing and increasing cost.

Public Safety Video Intelligence: a universal software platform you can customise for any AI situational awareness system. Operated securely either via the internet or locally, during live rapid chase and investigations of suspects. It automatically registers the identity of people entering the camera network and re-identifies them as they move throughout various locations where surveillance cameras are located:

The Command Control system is attached to the Viseum iVOS Global F3 system. This requires various input data such as types and locations of cameras, their respective transmission rates, and recording media, and the incident’s threat level and likely pursuit speed. It also determines the possible and likely travelling speed of the suspect.

When the suspect is identified either manually or automatically by Viseum iVOS software, the Global F3 system automatically follows them with alarms e.g. audio, text or visual display sent to command control, SMS text or email. When the suspect is hidden among similar-looking people, the surveillance operative is shown a gallery of the most likely matches for further examination.

This can also be used to automatically corroborate or proactively identify the location and cause of public incidents.

Caution when using inferior low-cost public safety AI situational awareness systems – as with using any low-cost system, its operational costs are higher and expected results are lower.

Demonstration history

2006 – Viseum iVOS F3 (Find, Fix and Follow) Track and Trace. This dates from when Viseum introduced coloured boxes to highlight people as a helpful way to introduce Video Analytics Software technology. This demonstration uses Viseum’s fixed colour contextual view camera, which is used to train and certify all Viseum iVOS software engineers on our core standards of video analytics software.

This demo automatically verifies and indexes its surveillance targets. It automatically gives each pedestrian a unique number to follow them through a city. This also shows it can ignore types of people if the site or security application requires.

Tag and Track Technology Concept – PTZ Camera Auto Tracking The moving PTZ Camera automatically verifies and indexes its surveillance targets.

2005 – Viseum’s Software Technology Service for High-Density Pedestrian Saliency This Demonstration uses a Viseum 360 Security Camera. This is detecting salient targets in high-density (pedestrian/traffic) areas. It automatically moves the PTZ Camera to follow the salient target in areas under surveillance.

A busy London shopping street viewed live over a low-bandwidth mobile phone data connection. The PTZ Camera’s track and trace images are also transmitted back to command control over a high-bandwidth fibre connection.

Typical pilot programme

Global F3 Software Technology Services are introduced in stages, starting with a pilot that has agreed success criteria. Typical durations are in working days.

Download the Global F3 Find, Fix and Follow brochure (PDF).

Frequently asked questions

Does Global F3 work with our existing cameras?

Yes. It is designed to work with analogue and digital cameras of any brand, operated locally or securely over the internet. Camera position and image quality affect results, so we assess them during the pilot.

Does the software decide who is a suspect?

No. It finds and ranks likely matches. A trained operator confirms each match and decides what to do.

Can it follow a vehicle as well as a person?

Yes. It follows vehicles across cameras, and with Vehicle Cross-Check it checks that the vehicle matches its registered plate.

How long does a pilot take?

Pilot design takes about 40 working days. Pilot deployment takes 110 to 200 working days, depending on the length of the trial agreed with you.

To see Global F3 find, fix and follow on your own cameras, book a private briefing or a live demonstration with the Viseum team.