Linker Vision ← All products / Campus

Campus

Campus Safety

Real-time campus safety monitoring across buildings, grounds, and entrances — detecting weapons, falls and medical incidents, fights, intrusions, loitering, fire/smoke, and unusual gatherings — flagged to security for human review.

VLM output

A daytime CCTV view of a campus tree-lined grounds separated by a low metal railing from an adjacent sidewalk and street, with a red 'Virtual Fence' zone drawn over the grounds. In the first frame a person in dark clothing is detected climbing over the railing at the boundary; later frames show the walkway clear and then a man in a red shirt walking along the sidewalk away from the camera.

  • Perimeter Intrusion DetectedYes
  • Virtual Fence CrossingYes
  • Fence Climbing BehaviorYes
  • Persons Detected1-2
  • Restricted Zone OccupiedNo
  • Loitering DetectedNo
  • Alert LevelHigh
  • Lighting ConditionDaylight
  • Crowd DensityLow

Customer challenges

1

Campuses are large, open, and busy, often with vulnerable populations; manual CCTV monitoring can't cover every camera, and incidents escalate before staff notice.

Customer benefits

24/7 proactive monitoring, faster response, safer students and staff, wide coverage of existing cameras, evidence capture, and reduced security workload.

Key features

Weapon Detection

Detects visible weapons in camera view and alerts campus security.

Incident Detection — Person Fall

Detects a person falling and raises an incident alert.

Fight / Aggression Detection

Detects fighting or aggressive physical behavior.

Unauthorized Access & Intrusion Detection

Detects entry into restricted or closed areas.

Loitering & Dwelling

Flags people lingering in an area longer than expected.

Fire & Smoke Detection

Detects fire and smoke early in monitored areas.

Crowd & Gathering Monitoring

Monitors crowd size and gatherings in campus spaces.

Before & after

Before Incidents caught by chance on monitors or reported after they escalate. After — with Linker AI Nexa Safety events flagged to security within seconds, with location, for human review.