Skip to main content

Architecture

VisualCortex is a highly scalable, Enterprise grade application.

The application itself can be run in the following configurations:

  • Utilising a Cloud Provider (AWS, GCP, Azure etc)
  • Utilising a Virtual or Physical Machine on-premises
  • On-premises in an Air-gapped environment

Concepts​

VisualCortex is made up of two main deployable components. These are:

  • The Control Plane, which houses the platform configuration and reporting interface
  • Edge Clusters

The platform interface has to be accessible by Edge Clusters and users wishing to interact with the platform. This typically means hosting this component in a central location accessible to users - the most common being a cloud tenancy.

Edge Clusters are more likely to be in more protected environments - either part of an existing CCTV network, or at least able to connect to any required live sources. Edge clusters also need to be able to communicate outbound to the platform configuration and reporting interface; for most customers, this is achieved with outbound internet access. Edge Clusters can successfully work behind NAT Gateways, through VPNs, and can also work with Corporate Proxies, provided that those proxies have support for Web sockets.

An overview of how different components interact in a Google Cloud environment is below. Differing cloud vendors utilise different components with similar functionality.

High Level Enterprise Architecture