# frigate_plate_recognizer

> Self-hosted Docker service that identifies license plates from Frigate events using Plate Recognizer or CodeProject.AI and publishes results via MQTT.

- **Category:** Video Surveillance
- **Licence:** MIT
- **Latest version:** v2.4.0
- **Open source:** oss
- **Languages:** Python
- **Platforms:** Docker, Docker Compose
- **Tags:** Mqtt, Home Assistant, Nvr, Frigate, License Plate Recognition, Docker, Self Hosted, Plate Recognizer, Codeproject AI, Vehicle Detection

## Description

Frigate Plate Recognizer is a self-hosted Docker service that enhances Frigate NVR by adding license plate recognition. It subscribes to Frigate MQTT events, sends vehicle snapshots to Plate Recognizer (cloud or self-hosted) or CodeProject.AI for license plate detection, and publishes enriched events with plate numbers back to Frigate via MQTT. The service supports watched plate lists with fuzzy matching, duplicate snapshot skipping for stationary vehicles, optional snapshot saving with plate bounding boxes, and Home Assistant sensor integration. Configuration is done via a YAML file. It requires Frigate, an MQTT broker, and either a Plate Recognizer API key (2,500 free lookups/month) or a local CodeProject.AI instance.

## Links

- [Homepage](https://github.com/ljmerza/frigate_plate_recognizer)
- [Docker](lmerza/frigate_plate_recognizer)
- [GitHub Repository](https://github.com/ljmerza/frigate_plate_recognizer)
- [Docker Hub](https://hub.docker.com/r/lmerza/frigate_plate_recognizer)

---
Source: https://selfhost.directory/project/frigateplaterecognizer