Skip to main content

Eye Detection

Introduction​

This section teaches how to use OpenCV to detect eyes in an image.

Experiment Objective​

Detect eyes in an image and draw rectangular boxes for display.

Experiment Explanation​

Using the cascade classifier method introduced earlier, this section uses the eye detection cascade classifier haarcascade_eye.xml. The code flow is as follows:


Reference Code​

The reference code is as follows:

'''
Experiment Name: Eye Detection
Experiment Platform: WalnutPi 1B
'''

import cv2

img = cv2.imread('eye.jpg') # Read the image

# Load the eye detection cascade classifier; note: the path must not contain Chinese characters
eyeCascade = cv2.CascadeClassifier('data/haarcascade_eye.xml')

# Detect all eyes
eyes = eyeCascade.detectMultiScale(img, 1.2)

# Iterate over all results
for (x, y, w, h) in eyes:
    cv2.rectangle(img, (x, y), (x+w, y+h), (0, 0, 255), 3) # Draw a box
    
cv2.imshow('result', img) # Display the image

cv2.waitKey() # Wait for any keyboard key to be pressed
cv2.destroyAllWindows() # Close the window

Experiment Results​

Run the above code on the WalnutPi; the experiment results are as follows:

eye_detection

Using a USB Camera for Recognition​

Combined with the USB camera usage method introduced earlier, you can perform real-time recognition via a USB camera. The reference code is as follows:

Reference Code​

'''
Experiment Name: Eye Detection (Using a USB Camera)
Experiment Platform: WalnutPi 1B
'''

import cv2, time

# Load the eye detection cascade classifier; note: the path must not contain Chinese characters
eyeCascade = cv2.CascadeClassifier('data/haarcascade_eye.xml')

cam = cv2.VideoCapture(1) # Open the USB camera

# Lowering the resolution can improve recognition speed; you can set it to 480×320 or 320×240
cam.set(3,480) # Set the captured image width to 480
cam.set(4,320) # Set the captured image height to 320
 
# Calculate FPS (frames per second parameter)
start = 0
end = 0

while True:
    
    start = time.time() # Record the start time
    
    retval, img = cam.read() # Read images from the camera in real time

    # Detect all eyes
    eyes = eyeCascade.detectMultiScale(img, 1.15)

    # Iterate over all results
    for (x, y, w, h) in eyes:
        cv2.rectangle(img, (x, y), (x+w, y+h), (0, 0, 255), 3) # Draw a box
        
            
    end = time.time() # Record the end time
    
    # Calculate FPS (frames per second), round to integer
    fps = round(1/(end-start))
    print('FPS: ', fps)
    
    # Write text on the image
    cv2.putText(img, "FPS: "+ str(fps), (20, 70), cv2.FONT_HERSHEY_SIMPLEX, 2, (0, 255, 0), 5)
    
    cv2.imshow('result', img) # Display the image
    
    key = cv2.waitKey(1) # The window image refresh interval is 1 millisecond to prevent blocking    
    if key == 32: # If the spacebar is pressed, break out
        break
    
cam.release() # Close the camera
cv2.destroyAllWindows() # Destroy the window displaying the camera video

Experiment Results​

eye_detection