fugleramme

Fugleramme is a Raspberry Pi project that turns a screen into a living bird picture frame. A microphone listens to the garden, a local BirdNET-Go instance identifies species by sound, and the frame renders the birds currently heard as hand-cut 19th-century natural-history illustrations. It is designed to run fully locally, with no cloud dependency for detection. How it works: BirdNET-Go performs the audio detection. Fugleramme polls its API, matches each detected species to an illustration, lays the cut-outs onto a textured paper page (larger birds toward the centre, sized by body mass), and redraws only when the set of birds changes. An admin page lets you configure what is shown. If you already run BirdNET-Go, the frame can point at it on the same machine or elsewhere on the network. Display options: the intended setup uses an Inky Impression 13.3" Spectra 6 e-ink panel in an A4 frame, but the panel is optional. Without one, Fugleramme runs web-only and the page adapts to whatever shows it - a TV, an HDMI display, another device on the network, or a desktop wallpaper or screensaver. Artwork: the project ships over 1000 cut-outs covering more than 500 species, each taken from a real public-domain plate and hand-curated. The README states no artwork is AI-generated, though some has been retouched with AI. Coverage is strongest in Europe and northern Asia, good in North America, and thinner in the tropics and southern hemisphere. An empty window shows a bare perch. Installation and deployment: a one-line install script for a Raspberry Pi clones the repo, installs dependencies, and runs the frame as a systemd service. A container image is also published, exposing a kiosk on port 8080 and an admin page at /admin, persisting data in /data. A Docker Compose example can bring up BirdNET-Go alongside it on a Linux box with a USB mic. For development, uv-based commands start a fake detector and a hot-reloading service. Licensing: code is MIT; the separately installed BirdNET-Go detection stack is CC BY-NC-SA 4.0 (non-commercial); the classic artwork style is CC BY-SA 4.0 with per-style attribution manifests; fonts are SIL OFL 1.1; bird body-mass data comes from AVONET under CC BY 4.0. The README notes the project is in early development and lists several similar projects it shares no code or art with.