Toggle Raspberry Pi GPIOs when a camera detects an authorized person.
Using face_recognition this script can toggle the GPIOs of a Raspberry Pi with pigpio.
Please note that the script will simply look for any faces in the video stream. It does not check if it is a real person or just a printed image of the face!
You can connect a camera via a rtsp stream or similar, or use a directly connected camera. The script will scan the current frame for faces and compare those to your defined reference images. If any face matches a reference face a GPIO of your pi will be turned on for a defined amount of time.
RaspiFaceID will automatically draw a box around the recognized faces and save the frame if an authorized person was found and the GPIO therefore turned on. If your reference image contain the persons name, e.g. Alex.png it will also display their names in the console and their name will be put under the drawn box in the frame.
This repository provides a default configuration, which will be copied to the project root folder. You need to edit this config.ini file to your needs.
Typically you have to configure at least:
- Video URL (Camera Data)
- RaspberryPi IP-Address
- GPIO address of the Pi that you want to toggle
The images will be stored inside the /data/images folder, which will also be created on the first startup. Note that you will have to rerun the container after you added new reference images or removed some.
- Python 3.3+
- macOS or Linux (Windows is also supported, but you will need to install
cmakeanddlibcorrectly) - Optional: RaspberryPi with pigpio installed
Make sure to install Visual Studio for C++ first.
To start the script simply use python main.py. Make sure that you have installed all dependencies before via pip install <Module> or pip install --no-cache-dir -r requirements.txt.
You can run this script inside a docker container. Therefore you will have to clone this repository:
git clone https://github.com/MikeNipkow/RaspiFaceID.git
After that you can build the docker image with the included dockerfile:
sudo docker build -t raspifaceid .
Run the container with:
sudo docker run -p 8080:8080 -v ${PWD}:/home/app/src raspifaceid