In recent decades, the number, extent, and quality of birds’ natural habitats in our towns and countryside have deteriorated. As a result, the role of the birdwatcher has become increasingly important.
Birdwatchers identify birds in their natural habitats by observing them and listening to their calls and songs.
Work similar to that of a birdwatcher is carried out by autonomous bioacoustic recorders, also known as autonomous recording units, or ARUs.
These devices record surrounding sounds and store them locally. The recordings can then be analyzed by birdwatchers and/or specialized software to detect bird calls.
Unfortunately, there are few birdwatchers and bioacoustic recorders in the Cilento National Park area. For this reason, RCR Innovative Labs decided to develop a device that combines these two functions as far as possible.
Bird Station is a smart birdhouse currently under development. It listens to and analyzes local bird calls directly on the device. Since it is also a birdhouse, it can provide shelter for birds. A small camera installed inside allows users to observe its surroundings and/or the birds that choose it as a nest in real time.
For obvious reasons, it cannot replace a birdwatcher. However, users can connect to it day and night to watch a live stream of the nearby woodland and listen to bird calls that have already been identified by species.
The key feature of this monitoring station is that anyone can listen to calls that have already been selected and catalogued by species and detection time, as shown in the web interface we developed:
The first prototype we built, although not particularly attractive, worked successfully: we detected several bird species, with a high level of call-recognition accuracy, in two different areas of the province of Salerno.
In the prototype, we used a Raspberry Pi 4, a small computer with relatively low power consumption, a compact USB microphone with a surprisingly effective foam windscreen, and a Raspberry Pi Camera Module 3 NoIR, which provides good night-time video quality. The recognition software is open source and is called BirdNET.
We also designed a small device that makes it possible to watch the live stream on a television. Using a small remote control, users can navigate a very simple graphical interface designed for older people.
