Detect and Blur Faces with a Simple Function — image analysis for beginners

Automatically anonymize your images, video’s or video-stream

Mike Huls

Jan 15, 2022 — 4 min read

This person is now totally anonymous (image by Library of Congress on Unsplash)

Detect and Blur Faces with a Simple Function — Image Analysis for Beginners

Automatically anonymize your images, videos, or video-stream

In this article, we’ll create a handy function that you can use to blur faces. With this function, you can automatically blur faces in an image file, video file, and even in-stream like when you’re recording yourself with a webcam.

First, we’ll discuss how faces are detected, then we’ll extract the coordinates of all faces in a frame and blur that area. Nothing too difficult; let’s code!

Series

This article is part of a series about OpenCV image processing. Check out the other articles:

Reading images, video’s, your screen, and the webcam
Detecting and blurring faces (📍 You are here!)
Destroying Duck Hunt with template matching: finding images in images
Creating a motion-detector
Detecting shapes without AI (under construction; coming soon)
Detecting and reading text from images (under construction; coming soon)

Dependencies and setup

In this example, we’ll read our screen but remember that this function that we’re creating can be used for reading webcams, images, and video files as well. If you want to use another source then check out this article to learn how this is done.
In our case we’ll need the following imports:pip install opencv-python pillow

Next, we need to download two files from this link. These are pre-trained models that we are going to use for detecting faces (frontal and profile).

Creating the face-blur function

In this part, we’ll create a tool that anonymizes an image. Let’s walk through the code:

In this piece of code, we create a class that we can use to detect and blur faces. First, we initialize the two files that we’ve downloaded previously. Now we have two classifiers that we can use to detect faces. We do this with the blur_face function.

Detecting faces

After one line of pre-processing on line 13, we can start to use our classifiers to detect faces. In lines 17 and 18 we analyze our gray frame to detect front-facing faces and faces that are slightly turned profile. We put both outputs in an array.

Blurring faces

The next step is to loop through all faces that we found. On line 22 we specify a region of interest; this is a snippet that contains only the face. In the next line, we blur this snippet. Lastly, in line 24, we paste this snippet back in its original place on the original image. Then we return the image.

Done! Let’s put this class to the test.

Using the face-blur function

In this part, we’ll use the function. If you’ve read this article you know you can pass anything to this function: frames from a webcam or video file, reading your screen, or even a static function. In this example, we’ll read a part of our screen and anonymize some images that I’ve found on Google.

There are just three steps:

Initialize class with the path to the cascades_folder (line 8)
Screenshot our screen at a specific location (line 12)
Pass the resulting image to the blur_face method on class (line 15)
Show the image (line 18)

That’s it! Lets see some results!

Anonymous people on unsplash (image by Michael Dam, Ryoji Iwata, Joseph Gonzalez, Rafaella Mendes Diniz and Ayo Ogunseinde)

Conclusion

In this part of the series we’ve handled image classifiers and modifying our image by blurring and overwriting to create a very easy-to-use function. We’ve seen that cascades are very easy to use and offer a lot of intelligence. In future articles we’ll examine how to train our own classifier; follow me to stay posted.

Don’t forget to check out the other articles in this series!

If you have suggestions/clarifications please comment so I can improve this article. In the meantime, check out my other articles on all kinds of programming-related topics like these:

Happy coding!

— Mike

P.S: like what I’m doing? Follow me!