Got it, let's work through this problem step by step. The core idea is to track how long the white object stays in a specific zone, trigger actions only after a 5-second delay, and reset everything when the object disappears. Here's a complete implementation that builds on your existing code:

Full Code Implementation #

import cv2
import time

# Initialize timer state variables
start_time = None
current_zone = None  # Can be 'upper', 'lower', or None
action_triggered = False  # Prevent repeated triggers after first activation

cap = cv2.VideoCapture(0)
cap.set(cv2.CAP_PROP_FRAME_WIDTH, 640)
cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 480)

def upper_zone_action():
    """Define your action for upper zone here"""
    print("Object stayed in upper zone for 5 seconds! Executing upper action...")
    # Add your actual operation here (e.g., send signal, control device)

def lower_zone_action():
    """Define your action for lower zone here"""
    print("Object stayed in lower zone for 5 seconds! Executing lower action...")
    # Add your actual operation here

while True:
    ret, frame = cap.read()
    if not ret:
        break

    # --- Your existing white object detection code ---
    # Convert to HSV for better color filtering (adjust ranges as needed)
    hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
    lower_white = (0, 0, 200)
    upper_white = (180, 25, 255)
    mask = cv2.inRange(hsv, lower_white, upper_white)
    contours, _ = cv2.findContours(mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

    moments = None
    if contours:
        largest_contour = max(contours, key=cv2.contourArea)
        moments = cv2.moments(largest_contour)

    # --- Timer and action logic ---
    if moments and moments['m00'] > 0:
        x = int(moments['m10'] / moments['m00'])
        y = int(moments['m01'] / moments['m00'])

        # Determine current zone
        if 0 < x <= 640 and 0 < y <= 240:
            zone = 'upper'
        elif 0 < x <= 640 and 240 < y <= 480:
            zone = 'lower'
        else:
            zone = None  # Object is outside valid frame area

        if zone:
            # First time detecting object in this zone
            if current_zone != zone:
                start_time = time.time()
                current_zone = zone
                action_triggered = False  # Reset trigger flag when zone changes

            # Check if 5 seconds have passed
            elapsed_time = time.time() - start_time
            if elapsed_time >= 5 and not action_triggered:
                if zone == 'upper':
                    upper_zone_action()
                else:
                    lower_zone_action()
                action_triggered = True  # Prevent repeated triggers

        # Draw debug info (optional)
        cv2.circle(frame, (x, y), 5, (0, 255, 0), -1)
        cv2.putText(frame, f"Zone: {zone}", (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)
    else:
        # No object detected: reset all timer states
        start_time = None
        current_zone = None
        action_triggered = False
        cv2.putText(frame, "No object detected", (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 255), 2)

    cv2.imshow('Frame', frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()

Key Details Explained #

• State Tracking: We use start_time, current_zone, and action_triggered to keep track of when the object entered a zone, which zone it's in, and whether we've already triggered the action for that zone.
• Zone Detection: We explicitly check if the object's centroid is within valid frame bounds (0 < x ≤640, 0 < y ≤480) before assigning a zone.
• Timer Logic: When the object enters a new zone, we reset the start time. If it stays in the same zone for 5 seconds and we haven't triggered the action yet, we run the corresponding function.
• Reset on Loss of Detection: As soon as no object is detected, we reset all state variables to prevent false triggers from leftover timer values.
• Debug Visualization: Optional circles and text help you verify the centroid position and zone status in real-time.

Customization Tips #

• Adjust the HSV ranges for lower_white and upper_white to match your specific white object's color (lighting conditions can affect this).
• Replace the upper_zone_action() and lower_zone_action() functions with your actual operations (e.g., controlling a motor, sending a network request).
• If you want to re-trigger the action after the first trigger (e.g., if the object stays in the zone for another 5 seconds), remove the action_triggered flag and logic.

内容的提问来源于stack exchange，提问作者Luisito