blakeblackshear/frigate

NVR with realtime local object detection for IP cameras

What it solves

Frigate NVR is a local Network Video Recorder (NVR) that provides real-time object detection for IP cameras. It solves the problem of generic video recording by adding AI-powered intelligence to surveillance footage, allowing users to identify specific objects of interest rather than relying on simple motion detection.

How it works

The system uses OpenCV and TensorFlow to perform object detection locally. To maximize performance and minimize resource usage, it employs a low-overhead motion detection system to trigger object detection only when and where it is necessary. It leverages multiprocessing to prioritize real-time processing over processing every single frame, and runs TensorFlow processes in separate processes for higher FPS. The system can be integrated with Home Assistant via a custom component and communicates via MQTT for broader system integration.

Who it’s for

Home automation enthusiasts and users with IP cameras who want a local, AI-powered surveillance system that integrates tightly with Home Assistant.

Highlights

  • Local AI Detection: Uses TensorFlow and OpenCV locally to avoid cloud dependency.
  • Resource Efficient: Only runs object detection on areas with motion, reducing CPU/GPU load.
  • Home Assistant Integration: Deep integration for smart home automation.
  • Flexible Recording: Supports 24/7 recording and object-based retention settings.
  • Low-Latency Viewing: Provides WebRTC and MSE support for live views.
  • AI Accelerator Support: Optimized for GPUs and AI accelerators to improve performance.

Related

  • Project
    blakeblackshear/frigate-hass-integration

    A Home Assistant integration that connects Frigate's AI-powered object detection and NVR capabilities to a smart home hub for automated security monitoring.

  • Project
    roflcoopter/viseron

    A self-hosted, local-only NVR and AI computer vision software for home and office monitoring with object and face recognition.

  • Project
    roryclear/clearcam

    An AI-powered NVR system that adds object detection, tracking, and AI-generated notification summaries to any RTSP security camera.

  • Project
    SharpAI/DeepCamera

    An open-source AI camera platform that enables local deployment of VLM scene analysis and object detection skills with autonomous, hardware-aware installation.

  • Project
    RizwanMunawar/yolov7-object-tracking

    A computer vision project that integrates YOLOv7 and YOLOv8 for real-time object detection and tracking across video streams from various sources.