Reference

Video surveillance glossary

120+ terms with crisp definitions, aliases, and cross-references to deep-dive articles. From VMS, NVR, and DVR to ONVIF profiles, RTSP, edge AI, LPR/ANPR, federation, and BIPA — every recorder, standard, analytic, storage, and privacy concept in one place. Search by name, filter by chapter, or jump straight to a letter.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
ALL
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

No terms match your filters. Try a different keyword or clear filters.



LPR / ANPR

License-plate / automatic number-plate recognition — reading vehicle plates from a camera feed for parking, access, and traffic use cases.

L

Metadata stream

A parallel data channel from the camera carrying detections, bounding boxes, and events alongside video; what ONVIF Profile M standardizes.

M



Motion detection

Detecting pixel change in a scene to trigger recording or alerts; the cheapest analytic and a major storage saver, but noisy.

M



Motion recording

Recording only when motion is detected; can cut storage by an order of magnitude but risks missing slow or off-frame events.

M

Multicast

Sending one camera stream to many viewers as a single network flow; saves bandwidth on large viewing deployments but needs network support (IGMP).

M

NAS (Network Attached Storage)

File storage attached over the network; affordable surveillance storage for small-to-mid systems, limited by network and file-protocol overhead at scale.

N

Natural-language video search

Querying footage in plain words («person in a red jacket near gate 3 last night»); the vision-language-model feature that defines AI-native VMS.

N



NPU (Neural Processing Unit)

A chip on a camera or edge box specialized for running AI models efficiently; what makes on-device analytics practical.

N



NVR (Network Video Recorder)

An appliance or software that records IP cameras over a network; narrower than a VMS, which adds multi-site, roles, analytics, and scale.

N

Object classification

Assigning a detected object its type — person, car, truck, bicycle, animal; what turns raw detections into filterable, searchable events.

O



Object detection

Finding and labelling objects (people, vehicles) in a camera frame; the foundational analytic that feeds tracking and search.

O

Object left behind

Detecting an item that appears in a scene and stays after the person who brought it leaves — bags in transit hubs being the classic case.

O

Object storage

Cloud-style storage addressing video as objects (S3 and compatibles); cheap, durable, and effectively unlimited — the natural warm/cold tier for footage.

O



Object tracking

Following a detected object across frames within a camera's view; the basis of line-crossing, dwell, and path analytics.

O



ONVIF

The open standard that lets cameras and VMS software from different vendors interoperate; defines profiles for streaming, recording, and analytics.

O

ONVIF conformance

A device's tested, declared support of an ONVIF profile, listed in the official conformant-products database; verify it before buying — marketing "ONVIF-compatible" is a weaker claim.

O

ONVIF Profile A

The ONVIF profile for access-control configuration — managing credentials, schedules, and access rules across vendors; pairs with Profile C door control.

O

ONVIF Profile C

The ONVIF profile for door control and event monitoring in access-control systems; lets a VMS talk to door controllers the way Profile S talks to cameras.

O



ONVIF Profile G

The ONVIF profile for edge recording and storage — searching and retrieving video recorded on the device.

O



ONVIF Profile M

The metadata-and-analytics ONVIF profile — a standard way for a camera to publish detected objects and events to the VMS.

O



ONVIF Profile S

The ONVIF profile for basic video streaming and PTZ control; the baseline most IP cameras and VMS support.

O



ONVIF Profile T

The advanced-streaming ONVIF profile — adds H.265, imaging settings, and analytics-metadata streaming over Profile S.

O

Open-platform VMS

A VMS built to host third-party cameras, analytics, and integrations through a public SDK (Milestone being the archetype), competing on ecosystem rather than lock-in.

O

People counting

Counting entries and exits with overhead cameras or stereo sensors; retail's core metric feeding conversion rates and staffing decisions.

P

Perimeter protection

Layered detection along a site boundary — thermal and visible cameras, line-crossing and intrusion analytics, often radar or fence sensors fused by the VMS.

P



PoE (Power over Ethernet)

Delivering electrical power to a camera over the same Ethernet cable that carries its video; simplifies IP-camera cabling.

P

PPE detection

Checking that people in a zone wear required protective equipment — helmets, vests, harnesses; a safety analytic, so misses matter more than false alarms.

P

Precision

Of all alerts the system raised, the share that were real events; high precision means few false alarms. Always read together with recall.

P

Privacy by design

Building privacy into the system from the architecture stage — minimal capture, masking, retention, access — rather than patching it on after launch; GDPR Art. 25.

P



Privacy zone

A masked region of a camera's view (a neighbor's window, a keypad) that is permanently blanked so it is never recorded.

P