This is a javascript web server that is hosted on a Raspberry Pi Zero 2 W with a cammera attached to it.
A pretrained neural model is combined with the live video feed from the Pi to recognise objects (particularly birds) and save their photos to the memory.
Later on their photos are displayed on the web page that can be accessed from any browser.
Here you will find a short list of requirements for this project.
- Raspberry Pi
- Raspberry Pi Camera
- Raspberry Pi OS Lite
First, copy this repository locally to your Pi.
Now you can install the required software:
./install.sh
- During the instalation you might encounter an error message saying that process was terminated due to lack of memory. This is a known issue and this means that you dont have anough RAM to build the library and you can try to fix this by creating a swap file as described here: https://askubuntu.com/a/796997.
If you are running a different OS to the one specified or want to experiment with this project then you should modify the build stack.
After everything is set up without any errors you can run the project locally:
node app.js
And open your browser window on: http://<pi-ip>:3000
.
Sometimes the child process running libcamera-detect
might not stop when the web server is stopped. To fix this you can run the following command:
sudo killall libcamera-detect
As an alternative you can run this service via systemctl
. Copy this repository to the /home/pi
directory and copy the file rpi_bird_watcher.service
to the /etc/systemd/system
directory. Then run the following commands:
sudo systemctl daemon-reload
sudo systemctl enable rpi_bird_watcher.service
sudo systemctl start rpi_bird_watcher.service
This project was developed by Markiian Benovskyi.
Favicon downloaded from Smashicons.
For bugs, questions and discussions please use the Github Issues.
GNU General Public License v3.0 provided in the LICENSE file. Attribution is required by the license, please read it for more details.