face-detection.py

import cv2

# LOAD CASCADES
face_cascade = cv2.CascadeClassifier('haarcascades/haarcascade_frontalface_default.xml')
eye_cascade = cv2.CascadeClassifier('haarcascades/haarcascade_eye.xml')


# SET UP THE VIDEO RECORDER
cap = cv2.VideoCapture(0)


# MODIFY HOW THE VIDEO RECORDER
while True:
    # this will capture the video frame-by-frame
    ret, img = cap.read()
    
    # I create a copy of the input image in grayscale mode
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    
    # read the input image to detect faces
    faces = face_cascade.detectMultiScale(gray, 1.3, 5)

    for (x,y,w,h) in faces:
        trc = (x,y)             # face rectangle's top-right corner
        blc = (x+w,y+h)         # face rectangle's bottom-left corner
        rectcolor = (255,0,0)   # blue color
        thick = 2               # line thickness
        
        # draw a rectangle whenever a face is detected
        cv2.rectangle(img, trc, blc, rectcolor, thick)
        
        # slice image area containing the face
        roi_gray = gray[y:y+h, x:x+w]
        roi_color = img[y:y+h, x:x+w]
        
        # read the face area and detect eyes
        eyes = eye_cascade.detectMultiScale(roi_gray)
        
        for (ex,ey,ew,eh) in eyes:
            trc = (ex,ey)           # eye rectangle's top-right corner
            blc = (ex+ew,ey+eh)     # eye rectangle's bottom-left corner
            rectcolor = (0,255,0)   # green color
            thick = 2               # line thickness
            
            #draw a rectangle whenever an eye is detected
            cv2.rectangle(roi_color, trc, blc, rectcolor, thick)

    cv2.imshow('Face Detector', img)
    k = cv2.waitKey(30) & 0xff
    if k == 27:
        break

cap.release()
cv2.destroyAllWindows()