opencv_face_detect.py

#!/usr/bin/python3

import cv2

from picamera2 import Picamera2

# Grab images as numpy arrays and leave everything else to OpenCV.

face_detector = cv2.CascadeClassifier("/usr/share/opencv4/haarcascades/haarcascade_frontalface_default.xml")
cv2.startWindowThread()

picam2 = Picamera2()
picam2.configure(picam2.create_preview_configuration(main={"format": 'XRGB8888', "size": (640, 480)}))
picam2.start()

while True:
    im = picam2.capture_array()

    grey = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY)
    faces = face_detector.detectMultiScale(grey, 1.1, 5)

    print("Number of faces detected:", len(faces))  # Add this line to check if faces are being detected

    for (x, y, w, h) in faces:
        cv2.rectangle(im, (x, y), (x + w, y + h), (0, 255, 0))
        # If a face is detected, print "Face detected!"
        print("Face detected!")
        # Close the camera preview and exit the loop
        picam2.stop()
        cv2.destroyAllWindows()
        exit()

    cv2.imshow("Camera", im)
    cv2.waitKey(1)
    print("Loop continues...")  # Add this line to check if the loop continues after a face is detected