如何在OpenCV中实现人脸检测时仅执行一次Python命令而非重复触发

The issue with your current code is that it checks face_count every frame and prints the message every single time the condition is met—so as long as a face is in the frame, it spams the print statement each loop iteration. To fix this, we need to track the previous state of face detection and only trigger the action when the state changes (from no face to face, or face to no face).

Here's how to adjust your code:

1. Add a state-tracking variable (we'll call it last_face_detected) initialized to False (since we start with no face detected by default).
2. In each loop, determine the current state (current_face_detected = face_count > 0).
3. Only print the message if the current state is different from the last state. Then update the state variable to match the current state.

Modified Code #

try:
    import cv2
    import os
    import time  # Don't forget to import time, it was missing in your original code!

    cap = cv2.VideoCapture(0)
    pTime = 0
    cascPath = os.path.dirname(cv2.__file__) + "/data/haarcascade_frontalface_default.xml"
    faceCascade = cv2.CascadeClassifier(cascPath)
    
    # State tracking variable: tracks if a face was detected in the LAST frame
    last_face_detected = False

    while True:
        success, img = cap.read()
        img = cv2.flip(img, 1)
        gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
        cTime = time.time()
        fps = 1 / (cTime - pTime)
        pTime = cTime
        face_count = 0

        cv2.putText(img, f'FPS:{int(fps)}', (20, 30), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)
        faces = faceCascade.detectMultiScale(
            gray,
            scaleFactor=1.1,
            minNeighbors=5,
            minSize=(30, 30),
            flags=cv2.CASCADE_SCALE_IMAGE
        )
        for (x, y, w, h) in faces:
            cv2.rectangle(img, (x, y), (x+w, y+h), (0, 255, 0), 2)
            face_count += 1
            cv2.putText(img, 'Face num '+str(face_count), (x-10, y-10), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (0, 0, 255), 2)
        
        cv2.putText(img, f'Faces Detected: {face_count}', (20, 70), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)

        # Determine current face detection state
        current_face_detected = face_count > 0

        # Only trigger action if state has changed
        if current_face_detected != last_face_detected:
            if current_face_detected:
                print("a face was detected")
            else:
                print("the face magically disappeared")
            # Update the state tracker to current state
            last_face_detected = current_face_detected

        cv2.imshow("Face Recognition", img)
        if cv2.waitKey(1) & 0xFF == ord(' '):
            cv2.destroyAllWindows()
            break
except KeyboardInterrupt:
    print("[KeyboardInterrupt] Exiting...")
    time.sleep(2)
    exit()

Key Changes: #

• Added import time (your original code used time.time() but didn't import it—this would throw an error!)
• Added last_face_detected to track the previous frame's detection state
• Replaced the direct if face_count >0 check with a state-change check: we only print when current_face_detected is different from last_face_detected
• Updated last_face_detected after printing to ensure we don't re-trigger the same message until the state changes again

This way, you'll only get one print when a face first enters the frame, and one print when the last face leaves the frame—no more spamming!

内容的提问来源于stack exchange，提问作者user17839070