RedNode Security

Recognition • Multi-Camera Logging

Camera: Not started

RedNode Secure: Camera Intelligence Overview

Anchored in the header, this overview mirrors the RedNode Secure console with multi-camera overlays, unknown face queues, object memory, and export-ready logs. It runs locally via getUserMedia() and keeps data in localStorage for fast, on-device response cycles.

Usage notes
RedNode Secure is designed for on-device workflows. Run on https (or localhost) for Live CV, and expect a short warm-up on first model load.
Mode: Demo Recognition: Off Storage: localStorage Tip: Use HTTPS for camera on mobile
Controls
Start cameras, toggle demo or recognition modes, and fit feeds to the workspace.
OFF
Face Logging
Register known faces and synchronize with storage.
Camera Feeds
Select cameras, open streams, show overlays, capture, annotate, or pop out.
Camera Control
Choose streams and (optionally) bind each slot to a real device.
12 fps
On iOS, only one active camera is typical. Start with Camera 0 and use HTTPS.
If you only have one webcam, select Camera 0. If multiple cameras exist, pick a device per slot. Some browsers/devices limit simultaneous streams.
Labeling & Recall Panel
Assign names, recall events, and manage tracked objects.
Unknown Face Labeling
Queue + preview + name assignment
Unknown Faces: 0
Demo Mode adds “unknown faces” randomly. In Live mode, face crops are queued automatically.
Custom Labeling
Rename selected object label
Custom labels persist in localStorage.
Tune Panel — Buzz Alerts
Trigger an internal buzz when tuned conditions are met.
Idle
No tuned alert configured.
Object Recall & History
Select object to recall its event history
Selected Log Details
No selection
Selected log preview
No preview available
Select a log row, tracked object, or unknown face to view details.
Storage & Sync
Choose local vs server storage, consent, and sync
Storage & Sync
Choose local vs server storage, consent, and sync
Local active
Default storage
Local keeps everything on-device. Cloud checks the server before commits.
Activation
Not checked
WS status
Connecting…
Feedback
Ready to save.
Commits require ADMIN_TOKEN plus GitHub vars (GITHUB_ENABLED, GITHUB_TOKEN, GITHUB_OWNER, GITHUB_REPO, GITHUB_BRANCH) set in Dockerfile.api or your deployment environment.
Server + Consent
Pick Cloud and consent to trigger the server check; when you click Activate Server you will see loading, error, or commit-ready feedback automatically.
Optional details
Server uploads assume consent for streamlined sync.
CV Status / Debug
Camera permission, TF backend, model status, and stream count.
Status: Camera: Not started • Models: Idle • COCO-SSD: Idle • Streams: 0 active
Tracked Objects
Objects seen in logs (click to load history)
Object Interaction Logs
Time • Object • Event • Person • Held Duration • Details
Time Object Event Person Held Details