Multi Camera Tracking (MCT)

MCT is an optional overlay of the Smart Corridor stack that tracks the physical movement of individuals across multiple cameras in real time using body detection rather than face recognition. While CIGS groups face detections into anonymous identities, MCT tracks where a person physically is and how they move — their trajectory, zone transitions, and crossing events. It ships in the mct/ folder of the deployment package and is started with the rest of the stack when enabled.

What it needs

• Eight additional services, pulled from the same Innovatrics registry as the corridor images, joining face-matcher-network.
• An embedded-camera feed — MCT consumes the pedestrian data produced on the cameras themselves (see edge streams), so the corridor cameras must be edge devices supported by Face Matcher.
• A per-site calibration that maps each camera's view onto the corridor floor plan; it is done once at installation and repeated when a camera is moved.
• The MCT Visualizer on port 8004, a browser view of the tracked positions used during calibration and for checking coverage.

Because of the camera requirement and the calibration effort, MCT is a deliberate addition rather than a default — plan it during site design, not after.

Role in the corridor

MCT feeds directly into CBS, which uses it as the primary source for zone entry/exit events, person count changes, and movement updates. The Officer Monitoring App floormap renders MCT positions. Deployments without MCT fall back to Face Matcher and CIGS data for the zone signals — the topic names and event structures are identical either way, so integrators require no changes when MCT is introduced.