AI and Video Analytics

Perimeter Intrusion and Virtual Fence Analytics

A perimeter alarm based only on motion detection can produce many false alarms from trees, light and weather. We combine virtual fence, line crossing and zone entry rules with person and vehicle classification; we aim to reduce false alarms with direction, duration, schedule and object size filters, and measure the result in the pilot. The alarm is linked to PTZ, video verification and the notification flow.

Virtual fenceLine crossingZone entryDirection filterDuration filterScheduleThermal + visualPTZ steering

Service definition

What is perimeter intrusion detection and a virtual fence, and who is it for?

A virtual fence is an analytics rule that raises an alarm for a person or vehicle that crosses a line or zone defined on the camera image in a given direction, or stays within it for a given time. It’s used to catch after-hours intrusions, unauthorized entry into critical zones and fence-climbing attempts as they happen.

01

Do animals trigger alarms?

Classification and size filters reduce animal-triggered alarms but don’t eliminate them. So the rate is measured and monitored.

02

Can a siren or lighting be triggered?

Yes; an audible warning or lighting can be triggered via camera or VMS outputs. Before setting it up, we assess the deterrent effect and the impact on neighbors.

03

Can existing cameras be used?

Yes, if the angle and image quality are sufficient. Success drops with cameras looking parallel to the perimeter; we measure this in the pilot.

Project stages

How do we proceed?

01 / STAGE

Risk analysis

Perimeter and response.

02 / STAGE

Design

Cameras and rules.

03 / STAGE

Pilot

Trials and measurement.

04 / STAGE

Commissioning

Alarm flow and training.

Video Surveillance

Spot the intrusion as it happens, not after the fact

Let’s talk about your perimeter and response setup; we’ll come back with a written proposal covering cameras, rules and the alarm flow.