Genetec KiwiVision™: Distributed Video Analytics Architecture

Genetec KiwiVision™ is a modular video analytics platform designed as a native plugin for the Security Center ecosystem. The solution automates security monitoring and enables organizations to shift from reactive security to proactive threat detection.

Architecture and Algorithms

KiwiVision™ leverages a distributed architecture featuring a central management server and compute nodes that analyze video streams using deep learning neural networks. The algorithms are trained to detect objects and behavioral patterns, ensuring high accuracy even in challenging lighting conditions or dense crowd environments.

The system supports horizontal scaling with load balancing and GPU acceleration via NVIDIA hardware. A single GPU-enabled server can process 2–3 times more HD streams compared to CPU-only computations.

Infrastructure Integration

KiwiVision™ integrates into the existing Security Center infrastructure using standard configuration tools without the need for additional servers or APIs.

The solution is fully compatible with over 30,000 camera models integrated into Security Center, with no hardware replacement required. Events natively work with automated response mechanisms to trigger actions: from displaying a camera on the video wall to activating sirens—without any programming. This approach simplifies the system architecture and reduces total cost of ownership.

Functional Modules

The platform architecture is built around four core modules covering the full range of video analytics tasks, from intrusion detection to privacy protection.

Security Video Analytics — the primary module for detecting security threats. It detects zone intrusions, monitors direction of movement, identifies unauthorized passage through turnstiles, tracks abandoned or missing objects, and logs vehicle stops in restricted areas.

People Counter — a module for counting people through virtual lines (entries and exits) and assessing crowd density in specific zones. It supports GPU acceleration for processing more streams and generates alerts when defined occupancy thresholds are exceeded.

Privacy Protector — a unique module for privacy protection. It generates two parallel video streams: a public one with masking of all moving objects, and a private one with the original footage. The private stream is accessible only to authorized personnel with appropriate permissions and access audit logs.

Camera Integrity Monitor — a module for camera integrity monitoring. It automatically detects sabotage (blocking, tampering), technical issues (defocus, overexposure), and provides status widgets for all cameras in the system. Included in the base license without limitations.

The modular approach allows integrators to select only the necessary functionality, optimizing cost and system complexity.

Technical Advantages

The key advantage of KiwiVision™ is its full unification with the Security Center ecosystem. This eliminates the need for separate analytics servers, API integrations, or additional applications. The platform uses the existing VMS infrastructure, significantly reducing total cost of ownership and simplifying project architecture.

Starting from Security Center 5.11, core modules are included in the base license for up to 300 cameras. A hardware resource calculator helps accurately define requirements during the design phase. The ability to add servers without downtime enables flexible scaling. A single point of support and familiar tools minimize deployment time.

The solution’s reliability is proven by hundreds of global deployments across enterprises, government agencies, and critical infrastructure organizations. Compliance with the UK’s NPSA security standards ensures system integrity for the most demanding applications. KiwiVision Privacy Protector has received the unique European Privacy Seal certification, making it the world’s first video analytics solution to achieve this recognition, enabling deployment even in jurisdictions with the strictest privacy regulations.

Deployment Scenarios

In urban security systems, KiwiVision™ automates monitoring of public areas. The platform detects fights, abandoned suspicious objects, or crowd formations without continuous operator supervision.

Crowd monitoring displays real-time occupancy of public squares during mass events. This helps prevent stampedes and redirect pedestrian flows.

In transportation hubs, cameras with Area Protection analytics replace physical perimeter sensors. They instantly detect intrusions onto runways or rail tracks and transmit the intruder’s coordinates via integration with the RSA Surveillance module — a restricted area security management system that consolidates various sensors on an electronic site map. People Counter optimizes transit network operations by counting passenger flow to manage train intervals.

Retail uses the People Counter module for business analytics: counting visitors enables conversion rate evaluation and staffing planning based on peak hours.

People Counter enables monitoring of semi-private spaces with masking. This ensures property security without violating customer privacy.

Industrial sites use KiwiVision™ for round-the-clock perimeter surveillance over large areas. Thermal cameras with analytics detect intruders even in darkness, while the Camera Integrity Monitor ensures there are no blind spots in the security system.

Conclusion

KiwiVision™ represents a comprehensive approach to video analytics through deep integration with existing VMS infrastructure, eliminating the architectural complexity of multi-vendor solutions. Deep learning neural networks deliver high detection accuracy in real-world conditions, while the modular structure allows functionality to scale according to the site’s specifics. Including core modules in the base Security Center license for up to 300 cameras makes the solution economically viable for a wide range of projects — from corporate offices to urban security systems, where rapid deployment and minimal operational costs are critical.