Video Processing Platform (VPP)
VPP is the biometric engine that powers all face detection, recognition, and watchlist matching in the corridor and eGate platform. It is developed and maintained by the Innovatrics Platforms team — it is not a BCSC product. The Smart Corridor & eGate Hub consumes VPP as a read-only upstream source: it subscribes to VPP's event stream and proxies VPP's management data to the GUI, but never writes back to VPP.
VPP ships inside the Smart Corridors & e-Gates bundle, paired 1:1 with the Hub — it is not a separately operated shared service. Multi-node topologies are covered in Deployment Layouts.
What VPP owns
From the BCSC perspective, VPP owns four key entities: cameras (physical devices and stream configuration), watchlists (named subject collections such as Allowed or Denied), members / subjects (enrolled individuals with biometric templates), and identification history (a raw log of every identification attempt). These are surfaced to the GUI through the Hub's GraphQL proxy — integrators and operators never interact with VPP directly.
For full documentation on VPP — including camera management, watchlist configuration, enrollment APIs, and deployment — refer to the VPP documentation in Platform Docs.
VPP Adapter
The VPP Adapter is a Hub component that bridges VPP's raw event format to the Hub's unified event model. It connects to VPP's GraphQL subscription endpoint, determines the identification outcome from each event, and publishes normalized notifications to the biometric_events exchange. It contains no business logic — only translation and routing.
| Topic published | Trigger |
|---|---|
identification.match | Face matched against any watchlist |
identification.no_match | No match found in any watchlist |
identification.clearance.green | Match on an allowed watchlist |
identification.clearance.red | Match on a denied watchlist |
identification.condition_violation | Face obscured — identification not possible |
See the Event Catalog for full payload details on each topic.