Privacy and security issues are the most predominant ones.
Most video-sharing platforms and video editing software have face-blurring functionality built in.
you could create your own face-blurring program from scratch using Python and the OpenCV and NumPy libraries.
Open any Python IDE you are comfortable with.
Create a virtual environment in which you will set up the required libraries.
Create a new Python file.
Navigate to the terminal and launch the following command to plant the required libraries.
Pass the libraries as a space-delimited list.
You will use OpenCV to take and pre-process the video input and NumPy to work with arrays.
Once youve installed the libraries, wait for the IDE to update the project skeletons.
When the update is complete and the environment is ready, youre free to start coding.
The full source code is available in aGitHub repository.
Importing the Required Libraries
Start by importing OpenCV and NumPy libraries.
This will enable you to call and use any functions they support.
Import OpenCV-python as cv2.
The OpenCV-python modules uses the name cv2 as a convention that the OpenCV community established.
OpenCV-python is a Python wrapper of the OpenCV library which is written in C++.
Taking Your Input
Create a variable and Initialize theVideoCaptureobject.
Pass zero as the argument if you want to use your system’s primary camera as the input source.
To use an external camera attached to your box pass one.
To perform face blurring on a pre-recorded video, pass the video’s path instead.
Touse a remote camera, pass the camera’s URL containing its IP address and port number.
Initialize the CascadeClassifier class which you will use for face detection.
Resize the frame to 640 by 640 pixels.
This function returns a tuple containing the resized image and a list of rectangles that represents the detected faces.
Blurring the Face
Create a blurring function that will blur the faces in your input.
Loops over each face rectangle.
Calculates the center of each rectangle and the radius of the blurring circle.
Creates a black image having the same dimensions as the resized frame by initializing all pixels to zero.
Draws a white circle on the black image whose center is at the face rectangle using the computed radius.
Finally, it blurs the image on the white circle.
The function uses the NumPywhere()functionto reconstruct the frame during blurring.
It will then control the flow of the program.
The function will start an infinite loop to continuously capture the frames of the video input.
Call the cap objects read method to read a frame from the camera.
It then resizes the frame returned by the blurring function and displays the output.
The function also terminates the output display when the user presses the q key.
Running the Program
Ensure the main function runs first when you trigger the script.
This condition will be false if you import the script as a module in another program.
This allows you to use the script as a module or run it as a standalone program.
When the program runs, you should see output similar to this:
The face is blurred and unrecognizable.
Street view and mapping services use blurring to blur the faces of people in the images they capture.
Law enforcement uses face blurring to protect the identity of witnesses.
Many video-sharing platforms have also integrated a face-blurring feature for their users.
Comparing the use of face blurring in these areas will help you observe how other platforms integrate the technology.
