What Are The Different Types Of Video Analytics Software?

What is Video Analytic Software?

Video analytics software sits between cameras and outcomes. It turns raw video into structured events that a monitoring team, a SOC, or an operations platform can act on.

The market is fragmented because “video analytics” covers multiple jobs: perimeter intrusion detection, forensic search, people counting, vehicle intelligence, safety compliance, and more.

It also spans multiple deployment models: analytics running on the camera itself, analytics running on a server next to the VMS, and analytics delivered via cloud.

This guide covers the main types of video analytics software, plus the common vendor groupings you will encounter.

TLDR

Video analytics software falls into several types: intrusion and perimeter analytics, object detection and tracking, behaviour analytics, vehicle analytics such as ANPR, forensic search and investigation platforms, and specialised analytics like thermal based classification. Products also differ by deployment (edge, server, cloud) and integration method (VMS native, third party plugins, open APIs). Choosing the right type depends on risk, environment, network constraints, and operational workflow.

What You Will Learn

• how the main categories of video analytics differ

• what each type is best at

• how edge analytics differs from server and cloud

• which vendor families typically serve each use case.

What Are the Types of Video Analytics Software?

Now, let’s look into the various software solutions that integrate with surveillance and intruder detection systems.

Each application is designed to fulfil specific objectives and requirements. All in all, learning about the types of video surveillance analytics software is key for those in charge of security.

Intrusion and Perimeter Analytics

Perimeter analytics is focused on early detection along boundary lines, sterile zones, and approach routes. The typical outputs are intrusion events, line crossing events, loitering events, and zone entry events, usually with metadata such as object type, confidence score, direction of travel, dwell time, and a snapshot or clip for verification.

Detection rule families

Common rule primitives include line crossing, region entry, region exit, fence line approach, dwell time thresholds, direction of travel constraints, and time of day schedules. Many modern packages also support “sterile zone” logic, which treats certain regions as never valid for human presence at defined times.

Where it runs and why it matters

Edge perimeter analytics (on camera) reduces bandwidth because the system can transmit event metadata and a short clip rather than streaming everything at full resolution. Server based perimeter analytics can run heavier models and fuse multiple camera views, but demands compute and stable network paths.

Thermal based perimeter detection is often deployed where lighting is poor or where visible cameras struggle. FLIR, for example, positions thermal cameras with onboard analytics as suitable for long fence lines and difficult conditions, including poor weather and darkness.

Object Detection, Classification, Tracking

This category is the engine room of modern analytics. Object detection identifies entities in each frame, classification labels them, and tracking keeps identity across frames. Outputs include bounding boxes, tracks, counts, and object attributes.

Primary functions

Object detect and classify people, vehicles, faces, and other classes the model was trained for. Hanwha describes its AI cameras as detecting and classifying people and vehicles in real time using deep learning algorithms.

Why object tracking matters

Tracking is what turns detection into actionable security. A single detection can be noise. A consistent track across a fence line is a likely intrusion. Tracking also supports PTZ handoff, where a detection on a fixed camera cues a PTZ to auto track the target.

Behaviour and Activity Analytics

Behaviour analytics attempts to infer intent or abnormality based on movement patterns rather than simple presence. In security, the main value is early warning and prioritisation.

Security oriented behaviour patterns

Loitering, repeated approach to an access point, reverse movement, lingering in a no stop zone, clustering in restricted zones, and abnormal dwell time. These are useful in car parks, service yards, access routes, and boundary approach zones.

Safety and compliance behaviours

Hard hat detection, high visibility clothing detection, restricted walkway detection, and fall detection appear in some vendor portfolios, mainly where safety is a primary KPI.

Vehicle Analytics and ANPR

Vehicle analytics includes detection and tracking of vehicles, direction analysis, and classification by vehicle type. ANPR, also called LPR, extracts licence plates and often includes make model recognition when supported.

Where ANPR fits

ANPR supports access control, audit trails, and investigation. In perimeter security, it also supports threat intelligence by identifying repeated vehicle visits, tailgating patterns, or suspicious approach behaviours.

Forensic Search And Investigation Platforms

Some analytics products are designed less for live alerting and more for compressing investigation time. They generate searchable indexes over recorded video: attributes, object types, time ranges, and zones.

BriefCam is a well known example of analytics positioned around rapid review, real time alerts, and data analysis modules.

Why this is different

Forensic platforms are not always the right tool for perimeter intrusion detection. They can support investigations and provide operational intelligence, but real time perimeter protection typically needs low latency alerting and strong false alarm resistance.

Platform Vendors Versus Feature Vendors

In procurement, you will typically see two vendor families.

Platform level providers

These are VMS and unified security platforms that ingest video, manage cameras, and provide analytics integration points. Milestone XProtect is often positioned as a VMS platform with a wide device ecosystem and analytics integrations.

Analytics specialists

Specialists provide advanced analytics that plug into VMS platforms or run as standalone. This includes companies focused on forensic search, perimeter detection, or specific vertical analytics.

Summary: The Types of Video Analytics Software

Video analytics software is best understood as categories of outcomes plus deployment architecture. Perimeter intrusion detection analytics is not the same as forensic search. Edge analytics is not the same as server analytics. The right selection depends on what events you need, how you verify them, where compute will run, and how the analytics integrates with your VMS and monitoring workflow