Summary
Add a complete video capture and query pipeline to the edge-ai accelerator. This introduces a new blueprint (video-capture-query), a new application component (520-video-query-api), ONVIF camera deployment scripts, and enhanced continuous recording capabilities in the media capture service (503).
Motivation
Industrial edge AI scenarios require continuous 24/7 video recording from cameras with the ability to query and retrieve historical video segments by camera, location, and timeframe. Data scientists need a REST API to request video clips (e.g., "get video from camera-01 on January 20th from 10:00 to 10:30") for analysis, model training, and incident review.
Scope
New Components
- 520-video-query-api — Azure Function providing REST endpoints for time-based video queries with SAS URL generation, optimized blob filtering, and MQTT-triggered capture
- video-capture-query blueprint — end-to-end Terraform blueprint orchestrating media capture, blob storage, and video query API with lifecycle management and architecture diagrams
Enhanced Components
- 503-media-capture-service — continuous recording mode with ACSA (Azure Connected Storage Account) sync, multi-trigger support, and video processing improvements
- 508-media-connector — updated docker-compose and mock RTSP camera configurations
- 111-assets — new ONVIF camera deployment scripts (Bicep and Terraform), PTZ profile discovery, and quick PTZ test utilities
Documentation
- ONVIF camera quickstart — step-by-step guide for deploying ONVIF cameras with Azure IoT Operations
- Architecture Decision Records — dual-component video architecture (ADR-006), continuous video capture with ACSA sync, edge video streaming and image capture
- Blueprint architecture diagrams — 6 drawio/mermaid diagrams covering high-level architecture, component integration, blob storage organization, data flow, network topology, and deployment sequence
Blueprint Integration
- full-single-node-cluster — video capture query variables, outputs, and
video-capture-query.tfvars.example
Components Affected
src/500-application/520-video-query-api (new)src/500-application/503-media-capture-service (enhanced)src/500-application/508-media-connector (updated)src/100-edge/111-assets/scripts (new ONVIF tooling)blueprints/video-capture-query (new)blueprints/full-single-node-cluster (integration)docs/getting-started/onvif-camera-quickstart.md (new)project-adrs/Accepted/006-dual-component-video-architecture.md (new)
Summary
Add a complete video capture and query pipeline to the edge-ai accelerator. This introduces a new blueprint (
video-capture-query), a new application component (520-video-query-api), ONVIF camera deployment scripts, and enhanced continuous recording capabilities in the media capture service (503).Motivation
Industrial edge AI scenarios require continuous 24/7 video recording from cameras with the ability to query and retrieve historical video segments by camera, location, and timeframe. Data scientists need a REST API to request video clips (e.g., "get video from camera-01 on January 20th from 10:00 to 10:30") for analysis, model training, and incident review.
Scope
New Components
Enhanced Components
Documentation
Blueprint Integration
video-capture-query.tfvars.exampleComponents Affected
src/500-application/520-video-query-api(new)src/500-application/503-media-capture-service(enhanced)src/500-application/508-media-connector(updated)src/100-edge/111-assets/scripts(new ONVIF tooling)blueprints/video-capture-query(new)blueprints/full-single-node-cluster(integration)docs/getting-started/onvif-camera-quickstart.md(new)project-adrs/Accepted/006-dual-component-video-architecture.md(new)