Surveillance that sees and alerts

Automatic detection of firearms (pistols, revolvers, rifles, shotguns) and bladed weapons (knives, machetes) in camera footage. Once the threat is confirmed, the system alerts the security control room within seconds — before the first shot or the approach.

How it works

1. 1A YOLOv8 detector trained for threats locates firearm, bladed weapon and grenade candidates, with its own confidence threshold for each type.
2. 2Each candidate is checked by a vision-language model (CLIP), which compares the crop with weapons and with objects that are often confused with them — guitars and other instruments, tools, phones, pipes, helmets and masks. Only what is also recognized as a weapon goes forward.
3. 3Skeleton estimation checks whether the weapon is in a person's hand (or close to their body). An object with no carrier — hanging on the wall, on a table — only triggers a "suspicious object" notice, with no threat alarm.
4. 4The threat is only confirmed when the held weapon persists over several consecutive frames, eliminating single-frame detections.
5. 5The alert with an image is sent to the control room, mobile app or VMS/alarm integration.

Benefits

• Shorter response time to armed incidents.
• Proactive monitoring, without relying on an operator watching every camera.
• Immediate evidence for law enforcement.

Face enrollment and recognition for access control, event check-in, attendance tracking and personalized service. An enrolled face becomes a numerical signature (embedding) compared in milliseconds with every face detected by the cameras.

How it works

1. 1The face is enrolled from a photo or the camera (name and profile).
2. 2The SFace network generates a 128-dimensional vector that represents the face.
3. 3Each detected face is compared with the database by cosine similarity.
4. 4Identified and unregistered people trigger different events (welcome × alert).

Benefits

• Access and check-in without queues, cards or badges.
• Personalized service — the salesperson knows who has arrived.
• Auditable access log.

Virtual fences drawn on the camera image outline hazardous areas (machine working radius, presses, overhead cranes, docks, energized areas) or restricted areas (storerooms, perimeters). When a person enters, an immediate alert is triggered — and it can sound sirens or activate the machine interlock.

How it works

1. 1The user draws the zone on the camera image (free-form polygon).
2. 2The detector identifies people and vehicles; the position of the feet determines presence in the zone.
3. 3Each entry generates an event with photo, duration and time.
4. 4Webhook/PLC integrations make it possible to stop machines or trigger signaling.

Benefits

• Prevention of serious and fatal accidents involving machines and vehicles.
• Set up in minutes, with no construction work or physical sensors.
• Evidence for investigating incidents and near misses.

Automatic license plate recognition (LPR) at the gate, with real-time access decisions: the plate read is compared against the list of allowed plates (residents, employees, suppliers) and blocked plates (former contractors, past incidents). Allowed vehicles are let in, blocked plates trigger an immediate alert to security and unregistered vehicles are treated as visitors — all logged with time and photo.

How it works

1. 1Vehicles are detected and tracked; a dedicated detector locates each one's license plate.
2. 2An OCR trained on plates from more than 65 countries reads the characters.
3. 3The reading is corrected using the local plate format (letters × digits by position, e.g. O↔0, I↔1, B↔8) and voted over several frames, eliminating single-frame errors.
4. 4Confirmed plate → list lookup (the same plate is matched across old and new plate formats) → allowed, blocked or visitor.
5. 5Each pass creates an entry in the access log with plate, time, decision, confidence and a photo of the vehicle.

Benefits

• Entry without stopping the vehicle, with no tag, card or intercom.
• Immediate alert when a blocked plate tries to enter.
• Auditable log of every entry (plate, time, photo and decision).
• Simple registration, with owner, notes and expiry date (e.g. temporary contractors).

Facial monitoring of drivers and operators to identify signs of fatigue and microsleep — eyes closed for a prolonged time, rising PERCLOS, frequent yawning — sounding alarms and notifying the control room before an accident happens.

How it works

1. 1The face mesh (468 points) is computed on every frame.
2. 2The EAR (Eye Aspect Ratio) measures eye openness with a per-person adaptive threshold.
3. 3PERCLOS indicates the percentage of time with eyes closed (NHTSA standard for drowsiness).
4. 4MAR detects yawns; the combination of signals determines alert, fatigued or drowsy.

Benefits

• Prevents accidents caused by drowsiness — up to 20% of serious road accidents.
• Immediate alerts at the workstation and in the monitoring center.
• Fatigue metrics by shift for working-hours management.