SquintPRO · Cross-camera tracking
Follow a person or vehicle across cameras, on a map of your site
On a large site, an incident rarely stays in one camera's view. SquintPRO's cross-camera tracking helps follow the same person or vehicle from one camera to the next, with the cameras placed on a map of the site (GIS). This page explains how teams use it, what it needs, where it stops, and the rules to set before anyone uses it.
Published 5 min read
The short answer
- What it does: follows a person or vehicle from camera to camera across the streams SquintPRO analyses, with each camera's position shown on a map of the site.
- Who uses it: HSE investigators reconstructing an incident or near miss, control-room teams, and operations teams following a vehicle's route on a large site.
- What it does not do: use face recognition, see through areas no camera covers, or replace the person who confirms each step.
- What it needs: cameras connected to SquintPRO, their positions on a site map, and clear rules for who may run a track and why.
Why following someone across cameras takes so long by hand
Recorders are organised by camera and time. To follow a truck across a site by hand, someone finds it on one camera, guesses which camera it should reach next, finds the right minute there, and repeats. Every guess is a chance to lose the trail. On a site with dozens of cameras every step adds time, and a wrong turn can easily go unnoticed.
How cross-camera tracking works in SquintPRO
- Start from a sighting. Tracking begins from the person or vehicle as seen on one camera, which your team can first find with a text description or a reference photo in SquintPRO's video search.
- Follow it across cameras. SquintPRO analyses many streams at once and follows the person or vehicle between cameras, so you see where it appeared next, on which camera and at what time. Tracking works for both people and vehicles, and it does not use face recognition.
- Read it on the site map. The cameras are placed on a map of the site, so the result reads as movement through the site rather than a list of camera names.
- Confirm each step. A reviewer checks each step before it is used in a report or an investigation, because similar-looking people or vehicles can appear on the same route.
Where teams use it
| Situation | What tracking gives you | Check first |
|---|---|---|
| Reconstructing a near miss involving a vehicle | The vehicle's path before and after the event, across cameras | Were the cameras along that route connected, and where are the gaps between their views? |
| A contractor vehicle that went somewhere it should not | Where it went after entering, in order, with times | Identical fleet vehicles may be confused; confirm each step with a distinguishing detail |
| An alert from a detection, such as entry into a restricted zone | How the person or vehicle reached the zone, and where it went afterwards | The footage must still be inside the deployment's retention period; uniform coveralls and helmets make people look alike, so check each step |
Where tracking stops
- Gaps in coverage. A person or vehicle that leaves the cameras' view cannot be followed through the gap. The track continues only when the subject is found again on another camera.
- Look-alikes. Identical vehicles, uniforms and protective equipment make subjects harder to tell apart, which is why each step is confirmed by a person.
- Poor views. Distance, night lighting, crowding and objects in the way all reduce what a camera can show.
- Footage that is no longer held. Tracking works on footage still inside the deployment's retention period.
- Proof of absence. A track that ends does not prove the subject went nowhere else. It shows where the connected cameras saw it.
Rules to set before anyone runs a track
Following a person across cameras is one of the most sensitive uses of video analytics. Settle these points with your legal, security and HSE owners before go-live:
Read more
- Who may run a track, and who may see the results.
- For which purposes, written down: for example a safety incident, an entry into a restricted area or a vehicle route.
- People, vehicles or both. Decide whether tracking people is in scope at all, or only vehicles.
- A record of each track: who ran it, when, and for which purpose, kept by the team that owns the process.
- Where processing runs. SquintPRO can run on your premises, where the footage does not leave the site, or in the cloud with 3, 6 or 12 months of retention chosen at setup.
- How the workforce is told. Say what is tracked and why. Teams that see monitoring as surveillance tend to work around it.
How to test it on your site
Test tracking on routes where you already know the answer before relying on it:
- Stage a few known movements once the cameras are connected, for example a vehicle driven from the main gate to an unloading bay, and a person walking a set route at night.
- Measure the result: whether the route is found from end to end, how many wrong candidates the reviewer had to discard, and where the trail was lost.
- Compare it with the manual way for the same routes.
- Agree in writing what result justifies wider use.
A first deployment covers one site, a defined set of cameras and success criteria written down before it starts. Ask for it to be priced separately from a rollout, so the decision to continue rests on the results.
Part of the SquintPRO platform
Cross-camera tracking works alongside SquintPRO's detections and video search on the same cameras. SquintPRO is Interkey's own platform, deployed in energy, ports, smart-city projects and Interkey's own factory. Interkey's team surveys the cameras, connects SquintPRO to your video management and alerting workflow, and tunes the deployment after go-live.
Questions buyers ask
Do we need new cameras for tracking?
Usually not. SquintPRO is designed to work with an existing CCTV estate. The survey shows which cameras can be used as they are, which need adjusting, and where the gaps between cameras are on the routes you care about.
Can the footage stay on our site?
Yes. In an on-premise deployment the analysis runs on your infrastructure and the footage does not leave the site.
Does tracking use face recognition?
No. SquintPRO's cross-camera tracking follows people and vehicles between cameras without face recognition, and a reviewer confirms each step.
How is tracking priced?
SquintPRO is priced by quote, according to the site, the number of cameras and the use cases in scope. Include tracking in your request, with the routes you need covered, so the quote reflects it.
How accurate is it?
It depends on your cameras, their positions and the gaps between them. Measure it on staged routes during the pilot: how often the full route is found and how many wrong candidates appear. We do not quote a general figure.
Next
Related
Next step
Discuss tracking on your site
Tell us about the site, roughly how many cameras it has, and what you need to follow: vehicles, people or both, and between which areas. We will tell you which routes the current cameras can cover, where the gaps are, and how a short test would measure it.
Or directly
+966-11-2180999 [email protected]Tawuniya Towers, North Tower, 7th Floor, King Fahad Highway, Olaya, P.O. Box 56835, Riyadh 11564, Saudi Arabia