Skip to content

README.dockerhub.md

Latest commit

 

History

History
67 lines (51 loc) · 1.92 KB

README.dockerhub.md

File metadata and controls

67 lines (51 loc) · 1.92 KB

driveway-monitor

Receive customizable, AI-powered notifications when someone arrives in your driveway.

driveway-monitor accepts an RTSP video stream and uses the YOLO11 model to track objects. When an object meets your notification criteria, it notifies you via Ntfy with a snapshot image.

Images

  • cdzombak/driveway-monitor:VERSION-amd64: For amd64 hosts (supports CUDA and CPU; auto-detects at runtime)
  • cdzombak/driveway-monitor:VERSION-arm64: For arm64 hosts (Raspberry Pi 4/5, etc.)

Replace VERSION with a specific version (e.g., 1.0.0) or 1 for the latest v1.x release.

Quick Start

docker run --rm \
  -v ./config.json:/config.json:ro \
  cdzombak/driveway-monitor:1-amd64 \
  --config /config.json \
  --video "rtsp://192.168.0.77:7441/your-stream"

Docker Compose (with NVIDIA GPU)

services:
  driveway-monitor:
    image: cdzombak/driveway-monitor:1-amd64
    volumes:
      - ./config.json:/config.json:ro
      - ./enrichment-prompts:/enrichment-prompts:ro # optional, for AI enrichment
    command:
      [
        "--config",
        "/config.json",
        "--video",
        "rtsp://192.168.0.77:7441/your-stream",
      ]
    ports:
      - 5550:5550
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              capabilities: [gpu]
              count: all
    restart: always

Configuration

See the GitHub repository for complete configuration documentation, including:

  • Notification criteria and filtering
  • CEL expressions for advanced track filtering
  • Ollama/OpenAI enrichment setup
  • Health monitoring integration

License

GNU GPL v3. See LICENSE.

Author

Chris Dzombak · GitHub