Rishabh kumar's Blog

Real-Time Face Detection and Blurring with OpenCV: A Tutorial

Rishabh kumar's photo
Rishabh kumar
·

3 min read

In the realm of computer vision, OpenCV emerges as a powerful ally, offering a plethora of tools for a wide range of tasks. In this tutorial, we embark on an exciting journey into the world of real-time face detection and blurring using OpenCV. This innovative application of computer vision technology opens up intriguing possibilities.

Step 1: Installation of OpenCV

Before diving into the fascinating world of real-time face detection and blurring, it's essential to ensure that OpenCV is installed. Fear not; you can effortlessly achieve this with a simple pip command.

import cv2

Step 2: Loading the Haar Cascade Classifier

Our journey begins with the loading of a pre-trained face detection cascade classifier generously provided by OpenCV. This remarkable classifier, known as the Haar Cascade Classifier, stands as a reliable sentinel, proficient in identifying faces in both images and video frames.

# Load the pre-trained face cascade classifier
face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')

Step 3: Initializing Video Capture

With the classifier in our toolkit, the next stride involves initializing a video capture object. This object grants us access to the default camera, typically indexed at 0. Adjust this index if multiple cameras are at your disposal.

cap = cv2.VideoCapture(0)

Step 4: Detecting Faces

Now, the stage is set to detect faces within the live video stream. In this segment, we read a frame from the camera and initiate the face detection process.

status, frame = cap.read()

# Detect faces in the frame
faces = face_cascade.detectMultiScale(frame, scaleFactor=1.1, minNeighbors=5, minSize=(30, 30))

The detectMultiScale function astutely returns a list of rectangles, each representing a detected face with coordinates (x, y, width, height).

Step 5: Blurring Detected Faces

Intriguingly, for either privacy or artistic flair, we can opt to blur the detected faces. Here's the artistry behind it:

for (x, y, w, h) in faces:
    face = frame[y:y+h, x:x+w]
    blurr = cv2.GaussianBlur(face, (99, 99), 0)
    frame[y:y+h, x:x+w] = blurr

# Display the frame with face-blurring
cv2.imshow('Real-Time Face Detection and Blurring', frame)

This eloquent code dissects the face region, applies a Gaussian blur with finesse, and elegantly swaps the original face region with its smoothly blurred counterpart.

Step 6: Controlling the Loop

To gracefully exit the loop when your mission is accomplished, we introduce a simple keyboard event. In this example, the loop gracefully concludes when the Enter key (ASCII code 13) graces us with its presence:

if cv2.waitKey(2) == 13:  # 13 is the ASCII code for Enter, and 2 milliseconds is the latency       
       break

And with that, you've embarked on a captivating journey into the world of real-time face detection and blurring, masterfully harnessed through the power of OpenCV. 🎉

Unleash your creativity and explore the myriad possibilities this project offers. If you've found this tutorial helpful, consider sharing and giving it a thumbs-up! 👍😃

Complete code:

import cv2

# Load the pre-trained face cascade classifier
face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')

# Open the video capture device (0 for the default camera)
cap = cv2.VideoCapture(0)

while True:
    # Read a frame from the video stream
    status, frame = cap.read()

    # Detect faces in the frame
    faces = face_cascade.detectMultiScale(frame, scaleFactor=1.1, minNeighbors=5, minSize=(30, 30))

    # Apply Gaussian blur to the detected face(s) and keep the rest of the frame as it is
    for (x, y, w, h) in faces:
        face = frame[y:y+h, x:x+w]
        blurred_face = cv2.GaussianBlur(face, (99, 99), 0)
        frame[y:y+h, x:x+w] = blurred_face

    # Display the frame with blurred face(s)
    cv2.imshow('W', frame)

    # Break the loop when the Enter key is pressed
    if cv2.waitKey(2) == 13:  # 13 is the ASCII code for Enter
        break

# Release the video capture and close all OpenCV windows
cap.release()
cv2.destroyAllWindows()
Pythoncamera