This repo contains a docker image for providing the names of the birds that Frigate picks up
To use the docker images, you can run it with the following code:
docker run ghcr.io/ttopholm/frigate-birdnet:latest
| Variable | Description | Mandatory | Default Value |
|---|---|---|---|
| mqtt-broker | Mqtt host | X | |
| mqtt-port | Mqtt port | 1883 | |
| longitude | Longitude of the place where cameras are located | X | |
| latitude | Latitude of the place where cameras are located | X | |
| mqtt-reconnect-interval | reconnect interval for reconnecting to mqtt server is connection was lost (in seconds) | 5 | |
| subscribe-topic | topic for subscribing to Frigate events | frigate/+/audio/bird | |
| publish-topic | The topic where bird data is published to | birdnet/bird | |
| max-audio-duration | Max left of recording from Frigate (in seconds) | 60 | |
| max-analyze-workers | Max workers for the analyzer | 3 | |
| min-confidence | Minimum confidence for the result | 0.25 | |
| delete-frigate-event | Deletes event in Frigate (delete no matter outcome) | False |
{
"image_url":"https://cdn.download.ams.birds.cornell.edu/api/v1/asset/242173971/900",
"common_name":"Common wood pigeon",
"scientific_name":"Columba palumbus",
"start_time":1717265299.484562,
"end_time":1717265302.484562,
"species_code":"cowpig1"
"camera": "Backyard"
}
image_url and species_code will be per default empty, unless you add the ebird.org taxonomy file, see here. By using the integration, you can also get the common_name translated into your local language
If you want the common name, species code and the image url, you need provide a ebird taxanomy file with the image included. Per default the ebird.org api doesn't provide the image, but by using the ebird_generate.py it can generate the file.
If you want to use the file, you will also need to add a volume with the ebird_taxonomy.json file in.
docker run -v ./labels:/birdnet/labels ghcr.io/ttopholm/frigate-birdnet:latest
To use the tool, you need to install the dependencies in requirements-ebird.txt like this
pip install -r requirements-ebird.txt
Go to https://secure.birds.cornell.edu/cassso/login and create an user and obtain the api token.
To run the generation tool, run the follow:
python ebird_generate.py --ebird-api-token=[api token from ebird] --country=[country you are located in] --locale=[Your locale]
country can be either your ISO 2 letter country, or it can also include sub region, fx. DK-85 (Which is Denmark, Zealand)
locale is you locale, fx. da
If you want it in Home Assistant as an image use the follow MQTT Image code:
mqtt
image:
- name: birdnet
unique_id: birdnet
url_topic: "birdnet/bird"
url_template: "{{ value_json.image_url }}"
json_attributes_topic: "birdnet/bird"
json_attributes_template: >
{ "name": {{value_json.common_name}} }