Facial Recognition Module - Step by Step Guide



  • Hi all,

    Due to there being no guide on this and I was looking for one… I have decided to just make one…
    I only start figuring it out last night and so I am still setting up. So I will complete this post as I progress.

    I am using the following hardware

    • Raspberry Pi 3

    • Logitech 1080 HD USB Webcam

    • MagicMirror is installed and confirmed functioning correctly.

    The software I am using to do this is Mobaxterm.
    It might be easier to do all of this on the PI directly.

    The steps I have taken to get this working:

    Set up the tools

    1. Facial Recognition Tools

    on the github page click download as zip
    extract zip and remove -master from the end of the folder. Note: its the internal folder that you want with the same name.
    copy to MagicMirror/modules using your FTP or whatever folder browser you are using.

    run this:

    sudo apt-get update
    sudo apt-get install libopencv-dev python-opencv
    sudo npm install python-shell
    sudo apt-get install python-pip
    

    change to the tools module folder

    cd /MagicMirror/modules/MMM-Facial-Recognition-Tools
    

    run

    sudo npm install
    
    pip install -r requirements.txt
    

    Capturing training images

    Run

    python capture.py
    

    Decide whether you want to capture images from your web cam or convert existing .jpg images.
    Enter the name of the person you are about to capture. Images will be stored in a folder named after the captured person in training_data/.
    Follow screen instructions.

    Training model


    Make sure you have captured all your images.
    Run

    python train.py
    

    The script will automatically scan the directory training_data/ for your images.
    Wait. You will end up with a training.xml file in the current directory.
    Copy down the [‘name’, ‘name2’,‘name3’] part because you will later need it for setting up your mirror’s face recognition and to test your face recognition model.

    1. Test your steps so far

    Make sure your training.xml from running train.py is in this directory - /MagicMirror/modules/MMM-Facial-Recognition-Tools

    specify the face recognition algorithm in the environment with

    export FACE_ALGORITHM=1
    

    specify your user labels in the environment with

    export FACE_USERS=Alice,Bob,Casey,Doug
    

    Run

    python facerecognition.py.
    

    Success?

    1. Set up the module for the mirror
      on the github page for the MMM-Facial-Recognition page, click download as zip
      extract zip and remove -master from the end of the folder. Note: its the internal folder that you want with the same name.
      copy to MagicMirror/modules using your FTP or whatever folder browser you are using.

    Visit the terminal window and go to your newly create folder - /MagicMirror/Modules/MMM-Facial-Recognition and run

    sudo npm install
    

    Open your config.js file and add:

    {
    	module: 'MMM-Facial-Recognition',
    	config: {
    		// 1=LBPH | 2=Fisher | 3=Eigen
    		recognitionAlgorithm: 1,
    		// Threshold for the confidence of a recognized face before it's considered a
    		// positive match.  Confidence values below this threshold will be considered
    		// a positive match because the lower the confidence value, or distance, the
    		// more confident the algorithm is that the face was correctly detected.
    		lbphThreshold: 50,
    		fisherThreshold: 250,
    		eigenThreshold: 3000,
    		// force the use of a usb webcam on raspberry pi (on other platforms this is always true automatically)
    		useUSBCam: false,
    		// Path to your training xml
    		trainingFile: 'modules/MMM-Facial-Recognition/training.xml',
    		// recognition intervall in seconds (smaller number = faster but CPU intens!)
    		interval: 2,
    		// Logout delay after last recognition so that a user does not get instantly logged out if he turns away from the mirror for a few seconds
    		logoutDelay: 15,
    		// Array with usernames (copy and paste from training script)
    		users: [],
    		//Module set used for strangers and if no user is detected
    		defaultClass: "default",
    		//Set of modules which should be shown for every user
    		everyoneClass: "everyone",
    		// Boolean to toggle welcomeMessage
    		welcomeMessage: true
    	}
    },
    

    Change the line to the actual location. It’s probably still under the Tools Module

    trainingFile: 'modules/MMM-Facial-Recognition/training.xml',
    

    for me, I changed it to

    trainingFile: 'modules/MMM-Facial-Recognition-Tools/training.xml',
    

    Also, change the line called Users to add you list of users that you set up during the Capture.py section above.

    users: ['Joe'],
    

    Finally, change the line, if applicable

    useUSBCam: True,
    

    FINAL STEP!

    For each module in your config.js, set the profile under which it should load. E.g -
    {
    	module: 'example_module',
    	position: 'top_left',
    	//Set your classes here seperated by a space.
    	//Shown for all users
    	classes: 'default everyone',
    },
    {
    	module: 'example_module2',
    	position: 'top_left',
    	//Set your classes here seperated by a space.
    	//Shown for all users
    	classes: 'Joe',
    },
    

    So now the first example module will load for everyone and the second will only load for Joe.


    THIS IS AS FAR AS I AM .. I AM DOCUMENTING AS I GO

    @paviro
    Module Developer - Please feel free to copy this or use as you see fit. I hope it is helpful to anyone else



  • reserved


Log in to reply
 

Looks like your connection to MagicMirror Forum was lost, please wait while we try to reconnect.