BirdNET with Raspberry Pi
BirdNET-Pi is a real-time acoustic bird classification system designed for Raspberry Pi. It uses a USB sound card to pick up bird sounds, and classifies them locally using a pre-trained machine learning model. Audio and visual data is shared around the world. This was an open source demo project that I just had to check out so I did setup the platform according to the very well written instruction at comprehensive installation guide .
Some learnings from this project:
- The uSD card for Firmware Image with the code did not show up as 32GB in Mac environment but only 32Mbyte. After some googling I found a ”workaround” in https://thomas.vanhoutte.be/miniblog/reclaim-the-full-storage-capacity-of-an-sd-card-on-macos/ I thought, but it did not help fully. Some more googling and a suggestion to put SD card in Windows computer and use Filesystem tool there to redo the partitioning and after that I could rerun the reclaim commands and proceed with installation, but I also ended up mechanically breaking one small uSD card when putting it into the adaptor for large SD card reader (lessons learnt, alway have more uSD card that you need at home)
- The raspberry pi platform got really hot, most likely due to intense traffic over Wifi and processing of real time audio in FFT algorithms.
- The raspberry pi platform even if put in a box is not water/dust tight so maybe a 3D print box would be useful for this project. There were some 3D print drawing of a bird provided together with the demo.
- If you would put this device in a recording spot not close to your wifi network it could be fun to build a variant that collects and send over cellular network. There are plenty of possible devices to hook up to this platform for further development, but the data cost could be significant.
The above project is an open source project based on raspberry pi as HW platform, but there exist similar solutions that are more ”professional” and available as Android and iOS Apps for mobile phones https://merlin.allaboutbirds.org/the-story/
