Glossary
What is video analytics?
Definition
Video analytics is software that automatically analyzes video to detect and classify people, vehicles, objects and events, such as intrusion, loitering or missing safety gear. Modern AI video analytics uses computer vision models to process live streams in real time and raise alerts, so operators respond to events instead of watching every screen.
Video analytics explained
Earlier analytics relied on pixel-change rules such as motion detection, which react to shadows, rain and lighting changes. AI video analytics uses trained deep-learning models that recognize what is in the scene, which allows more specific rules.
Analytics can run inside the camera, on an on-site server or appliance, or in the cloud. Where it runs affects bandwidth, privacy and how many detections can run at once.
OrionAlerts
How OrionAlerts handles it
In OrionAlerts, analytics run on the same video pipeline that records the cameras: the AI runtime takes decoded frames from the Nebula OS video engine, so there is no parallel stream into a separate analytics box. Zone and schedule rules decide when an event raises an alert, and alerts reach operators on the dashboard, mobile app and email.
Explore AI Vision & Real-Time AlertsFAQ
Video analytics: common questions
Does video analytics replace security operators?
No. Video analytics watches every camera continuously and flags defined events, but people still verify alerts, decide on a response and handle situations the rules do not cover. The practical benefit is that operators spend their time on events that matter instead of scanning screens where nothing is happening.
Can video analytics work with existing CCTV cameras?
Often, yes. Analytics that run on a server or appliance can process streams from existing IP cameras over RTSP or ONVIF, and some setups can use streams from recorders. Results depend on camera placement, lighting, resolution and how large the target appears, so each view should be checked before relying on it.
Which detections does OrionAlerts support?
The detections OrionAlerts lists cover safety (PPE compliance, falls, fire and smoke), security (intrusion, line crossing, restricted-area entry, loitering, abandoned objects), operations (crowds, occupancy, vehicles) and camera integrity (tampering). Which ones apply depends on the site and each camera view, and custom use cases can be scoped during solution design.
Related terms
Edge AI video analytics
AI video analytics that run close to the cameras, in the camera or on an on-site server, instead of in a remote cloud.
Read morePPE detection
AI video analytics that check whether people in a zone are wearing required protective equipment such as helmets or safety vests.
Read moreCamera tampering detection
Analytics that detect when a camera is blocked, covered, moved, defocused or otherwise stops showing its intended scene.
Read moreANPR (automatic number plate recognition)
Video analytics that read vehicle number plates from camera images and turn them into searchable text for access control and logging.
Read more
See it on your own cameras
Book a walkthrough of OrionAlerts on Nebula OS, from camera connection to alerts on the dashboard, mobile app and email.