Detection

Loitering & Suspicious Behaviour

Identify individuals exhibiting unusual dwelling patterns, repeated passes through sensitive areas, and abnormal presence in restricted or low-traffic zones.

How It Works

Under the hood

The loitering model establishes baseline activity patterns for each camera zone. It learns what normal looks like: how long people typically spend in an area, which direction they move, and what times the zone is active. Anything that deviates significantly from this baseline triggers analysis.

When an individual remains in a zone beyond the expected dwell time, the system begins tracking. It measures duration, tracks whether the person appears to be waiting for someone specific or observing a location, and checks for repeated appearances in the same zone across different time windows. Someone who passes the same ATM three times in an hour generates a different alert than someone sitting on a bench.

The model also identifies reconnaissance-style behaviour: photographing security features, testing door handles, or repeatedly checking sightlines. These patterns, combined with dwelling time and location sensitivity, build a comprehensive risk picture that security teams can act on.

Why It Matters

The real-world impact

Most crimes are preceded by planning. Whether it's a theft, a targeted attack, or antisocial behaviour, the perpetrator typically scouts the location first. Loitering detection catches that preparation phase.

For retail, it identifies potential shoplifters casing a store. For transport hubs, it flags individuals who may be planning disruption. For public spaces, it helps identify antisocial patterns before they escalate. The key value is intervening during planning, not during execution.

48hrs

Baseline Learning

95%

Dwell Accuracy

30 days

Pattern Memory

Capabilities

What it detects

Purpose-built detection models tuned for real-world conditions and operational environments.

Adaptive Dwell Time Analysis

Learns normal presence duration for each zone and flags when individuals exceed expected times, adjusted for time of day and zone purpose.

Repeat Visit Detection

Tracks when the same behavioural signature appears in a zone multiple times across hours or days, identifying surveillance and casing patterns.

Reconnaissance Behaviour Flagging

Identifies actions consistent with location scouting, including testing access points, photographing infrastructure, and checking sightlines.

Zone Sensitivity Scoring

Different zones carry different risk weights. Loitering near a cash machine, emergency exit, or loading bay generates higher-priority alerts than a public seating area.

See it in action

Book a demo to see how loitering & suspicious behaviour works with your existing cameras. No new hardware required.

Free consultation. Works with any CCTV system. Live in under 48 hours.