Frequently Asked Questions About AI Kameram

AI Kameram is a video analytics solution that applies artificial intelligence to the footage of security cameras a business already operates. It detects flame, smoke, people and vehicles in the camera image and, on fire or unauthorised entry, raises an alert within seconds through an automated phone call, SMS, e-mail, mobile push notification and a dedicated alarm device in the guardhouse. Analysis runs on a GPU server installed inside the facility.

Yes. It works with any IP camera supporting ONVIF or RTSP. Analogue cameras can be connected through an NVR or DVR that exposes an RTSP stream. Brand does not matter: Hikvision, Dahua, Bosch, Axis and equivalents are supported. If you have no cameras or need to renew them, camera supply is also arranged.

Live analysis runs on a GPU edge server installed inside your facility. The camera stream never leaves the site. Only the evidence frame and event record of a confirmed alarm are sent to the central management panel; there is no continuous video upload. This preserves your internet bandwidth and keeps outbound data to a minimum.

The system is designed around processing personal data in place: footage is analysed inside the facility and only limited alarm evidence leaves the site. The panel provides per-user authorisation, access logging and retention management.

Note that lawful CCTV monitoring also requires operator-side duties such as privacy notices and camera signage; guidance is provided during installation.

Once flame or smoke enters the camera's field of view, detection and verification typically complete within seconds and the alert channels fire. Every camera is analysed continuously, and each detection is followed by a second verification stage confirming the event is real.

Yes. The fire model is trained to recognise both flame and smoke. In smouldering fires smoke becomes visible before flame, which makes it critical for early warning.

A smoke detector waits for smoke to rise and physically reach the sensor, so in open areas, high-ceilinged warehouses, greenhouses and outdoor yards it either responds far too late or never triggers.

AI Kameram looks at the image: flame or smoke is detected the moment the camera sees it. You also receive evidence imagery showing which camera and which spot — not just that there is a fire, but where. See the fire detection page for detail.

A detection never becomes an alarm on its own. In sequence:

• Physical filters — object size, aspect ratio and texture variance
• Static light filter — lamps, LEDs and reflectors glowing at the same spot are learned and removed
• Persistence and growth checks — real flame flickers and grows; single-frame flashes are ignored
• Day/night scene mode — night-specific thresholds engage when the camera turns black-and-white
• Visual verification layer — the event is confirmed a second time before alarming

In addition, per-camera zones and time schedules let you exclude known sources such as welding bays, stoves and exhausts during specific hours.

It works at night and thermal cameras are not required; standard IP cameras with IR night vision are sufficient. The system automatically recognises from the image that the camera has switched to black-and-white night mode and applies night-specific brightness, texture and shape criteria, so street lights, reflectors and vehicle headlights are not mistaken for flame.

For each camera you define which zone is monitored during which hours. An alarm is raised when a person is detected outside permitted hours or inside a restricted zone.

The system does not behave like a motion sensor; it checks whether there is genuinely a person, so swaying branches, animals, rain and moving shadows do not trigger alarms. Vehicle detection can be configured as a separate rule. See the intrusion detection page for detail.

Five channels fire simultaneously:

• Automated phone call — announces the event type and location
• SMS — short summary alert
• E-mail — with a screenshot of the moment, location, type and timestamp
• Mobile push — instant notification to panel users
• Guardhouse device — siren, flashing light and LCD display

Which camera notifies whom and during which hours is configured in the panel.

The dedicated device sounds a loud siren, flashes the fire or intruder icon in red and displays the type and location of the incident on its LCD screen, so the officer on duty is aware within seconds even without looking at a phone.

Because analysis runs on the on-site server, cameras keep being monitored during an internet outage and the local alarm device still works. Outbound notifications such as calls, SMS and e-mail are delivered once connectivity returns; event records are not lost.

A single on-site server can analyse dozens of cameras simultaneously depending on its GPU capacity, and hardware scales with camera count. Businesses with multiple facilities install one server per site and manage them all from a single central panel.

If your cameras are ready, installation typically completes within one working day. Requirements:

• Space and power for an on-site server on the same network as the cameras
• RTSP credentials for the cameras
• An internet connection so notifications can leave the site

Cameras are then registered in the panel, monitored zones are drawn and time rules are set.

Yes. AI Kameram does not replace your recording system; it adds an analysis layer on top. The NVR keeps recording while AI Kameram reads the same camera streams, analyses them and triggers the alert channels when an event occurs.

Yes. One on-site server is installed per facility and all sites are monitored from a single central panel. The panel supports organisation- and user-level authorisation: a site manager sees only their own facility while head office can see all sites.

Pricing depends on camera count, number of sites, the detection types required (fire, person, vehicle) and hardware needs. Request a demo for a concrete quote; your site is assessed and a tailored price is offered.

Yes. Fill in the demo form, write to info@aikameram.com, or reach us on +90 555 595 26 00 by phone or WhatsApp. The system is demonstrated live and the setup suited to your site is designed together.