This is my first big project so the code is a little messy! Also, I don't have the habit of commenting my code, so there is that.
The goal of this project is to automate tracking students by recognizing their faces. When a student enters to a class, the camera (a Raspberry Pi in this case) recognizes their faces and keeps a record of this in a database. Later we can find who attended the class by analyzing these records.
Currently, it is at a very early stage of development, so there are not many features.
We need at least 3 devices (a client, a camera and a server) to run this application. Server's purpose is to create a connection point between clients and cameras. Client application is where we'll connect to cameras and change their settings.
The steps below will take you through installing the application.
This is only tested in Linux, so if you are using MacOS or Windows it may not work properly.
-
Use Git to clone the repository or you can directly download it by using the 'Download ZIP' button.
$ git clone https://github.com/auriomalley/class-attendance-tracker.git -
Change directory into the newly created class-attendance-tracker folder.
$ cd class-attendance-tracker
I tested this with a Raspberry Pi model 4B 2GB RAM version with Raspbian installed. I used the lite version of Raspbian but it should work with other versions as well.
-
Install dependencies required for Pillow and OpenCV.
$ sudo apt install libjpeg-dev libatlas-base-dev -
Install Python dependencies. You can use Pip with a virtual environment, but I recommend doing it the Apt way since it also installs some needed libraries.
$ sudo apt install python3-pip python3-picamera python3-opencv opencv-data python3-pymongoIf you choose to do it using a virtual environment and Pip you may have to hunt down all other needed libraries on your own.
-
Install the libraries that are not available in the repositories using Pip.
$ sudo pip3 install pytz face_recognition tinydb -
Run the application.
$ python3 run_rpi.py -
This first run will generate a config.json file in the data/rpi directory. Go there and change the config settings.
If there are any null values in the config file, application will raise an error.
Client will allow us to connect to the main server and the database, so we can add students and receive reports. Installation is pretty straightforward.
Client application sometimes crashes. If that happens just relaunch it.
-
Install needed dependencies.
$ pip3 install pymongo pytz pillowI recommend using a virtual environment since there are no packages that would cause trouble later.
-
Run the application.
$ python3 run_client.py -
This first run will generate a config.json file in the data/client directory. Go there and change the config settings.
If there are any null values in the config file, application will raise an error.
Server allows us to connect the clients to the Raspberry Pis.
-
Install needed dependencies.
$ pip3 install tinydb pytzI recommend using a virtual environment since there are no packages that would cause trouble later.
-
Run the application.
$ python3 run_server.py -
This first run will generate a config.json file in the data/server directory. Go there and change the config settings.
If there are any null values in the config file, application will raise an error.
-
This application also requires a MongoDB database. I recommend using docker to install the MongoDB server. If you don't have it you can install it by following this documentation: Docker CE Installation.
$ docker pull mongo
$ docker run -d --network host -e MONGO_INITDB_ROOT_USERNAME=username -e MONGO_INITDB_ROOT_PASSWORD=password mongoThis will run the database.
After you installed the application, you can start by running the server. Raspberry Pi doesn't run the code on startup so make sure you have that set up, otherwise you'll have to connect to the Pi through Ssh to start the application every time. There is a tutorial on how to set it up: Run command on boot. Now you can run the client application and start adding students.