Faceblur

A simple python script to detect and blur faces in images. Made using face_recognition for the heavy lift and PyQt5 for the GUI

For better usability the script can be turned into a desktop app using Fbs, the only prerequisite is Python 3.5 or 3.6. Python 3.7 is not yet officially supported.

This project is meant as a design research tool to speed up the process of blurring faces and being GDPR compliant.

Setup

Open the terminal and clone this repo:

git clone https://github.com/tommasoltrz/FaceBlur.git

move into the repo directory:

cd FaceBlur

Create a virtual environment in the current directory:

python3 -m venv venv

Activate the virtual environment:

# On MacOS
source venv/bin/activate

Install the required libraries:

pip3 install fbs PyQt5==5.9.2 Pillow numpy face_recognition

Run the script

You can either directly run the script:

python3 ./src/main/python/main.py

or you can use Fbs (recommended):

fbs run

Turn it into an app

If everything works correctly with fbs run you can proceed to freeze the script into an app

fbs freeze

Test it using the command you'll get. for Mac:

target/Faceblur.app/Contents/MacOS/Faceblur

You can now see your app in the target folder

Troubleshooting

If you get the error unable to open .../your-path/shape_predictor_68_face_landmarks.dat simply copy and paste the folder face_recognition_models from your virtual environment into the app

cp -R .../your-path//FaceBlur/venv/lib/python3.6/site-packages/face_recognition_models .../your-path/FaceBlur/target/Faceblur.app/Contents/MacOS

Create installer

In order to make it easy to share you can create an installer by running

fbs installer

Credits

Most of this project was developed following these tutorials:

fbs-tutorial

PyQt5-tutorial