Eye Detector and Overlay

Face Detector with Eye Overlay Project

Overview

This project demonstrates an advanced use case of OpenCV (cv2) for detecting faces and eyes in real-time video streams. Beyond basic face detection, this project overlays custom images (from my portfolio) onto the detected eye regions, creating a playful and interactive effect. It’s a great example of combining computer vision techniques with creative visual effects.

Features

• Real-time face detection using Haar Cascade classifiers.
• Precise eye detection within the detected face region.
• Custom image overlay on detected eyes.
• Modular code structure for easy customization and integration.

Tools & Technologies Used

• Python 3.x
• OpenCV (cv2)
• Haar Cascade Classifiers:
  • haarcascade_frontalface_default.xml for face detection.
  • haarcascade_eye.xml for eye detection.
• Optional: matplotlib, numpy (for image processing/display).

How It Works

The project uses two pre-trained Haar Cascade classifiers to detect faces and eyes:

1. Face Detection:
   • The haarcascade_frontalface_default.xml classifier detects the entire face region.
   • The detected face region is used as a bounding box for further processing.
2. Eye Detection:
   • Within the detected face region, the haarcascade_eye.xml classifier identifies the eye areas.
   • These eye regions are then used to overlay custom images.
3. Image Overlay:
   • Predefined images from the portfolio are resized and placed over the detected eye regions.
   • The result is displayed in real-time via the webcam feed.

Code Snippet

Here’s a simplified version of the code that demonstrates the core functionality:

```python import cv2

Load the pre-trained classifiers

face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + ‘haarcascade_frontalface_default.xml’) eye_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + ‘haarcascade_eye.xml’)

Read the custom image to overlay on eyes

overlay_image = cv2.imread(‘portfolio_image.jpg’, cv2.IMREAD_UNCHANGED)

Start capturing video from the webcam

cap = cv2.VideoCapture(0)

while True: # Capture frame-by-frame ret, frame = cap.read()

# Convert to grayscale for detection
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

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

for (x, y, w, h) in faces:
    # Draw rectangle around the face
    cv2.rectangle(frame, (x, y), (x+w, y+h), (255, 0, 0), 2)

    # Region of Interest (ROI) for eyes within the face
    roi_gray = gray[y:y+h, x:x+w]
    roi_color = frame[y:y+h, x:x+w]

    # Detect eyes within the face ROI
    eyes = eye_cascade.detectMultiScale(roi_gray)

    for (ex, ey, ew, eh) in eyes:
        # Resize the overlay image to fit the eye region
        resized_overlay = cv2.resize(overlay_image, (ew, eh))

        # Overlay the image onto the detected eye region
        for i in range(ew):
            for j in range(eh):
                if resized_overlay[j, i, 3] > 0:  # Check alpha channel
                    roi_color[ey+j, ex+i] = resized_overlay[j, i]

# Display the resulting frame
cv2.imshow('Face Detector with Eye Overlay', frame)

# Exit loop when 'q' is pressed
if cv2.waitKey(1) & 0xFF == ord('q'):
    break

Release the capture and close windows

cap.release() cv2.destroyAllWindows()