Problem with counting people in the zone

[yavuzhantemiz]

Hi all,

I watched the tutorial for object counting in real time. I want to add minimum and maximum people limits. For example, when the maximum people limit reached, user gets a message about it. It works partly. It gives messages both within and outside the zone. I just want get messages when limit reached inside the zone i provided earlier. This is my code:

ZONE_POLYGON = np.array([
    [0, 0],
    [0.5, 0],
    [0.5, 1],
    [0, 1]
])

MIN_PERSON_COUNT = 3
MAX_PERSON_COUNT = 5

def parse_arguments() -> argparse.Namespace:
    parser = argparse.ArgumentParser(description="YOLOv8 live")
    parser.add_argument(
        "--webcam-resolution",
        default=[1280, 720],
        nargs=2,
        type=int
    )
    args = parser.parse_args()
    return args


def zone_feed():
    cap = cv2.VideoCapture(0)

    def generate():
        args = parse_arguments()
        frame_width, frame_height = args.webcam_resolution
        cap.set(cv2.CAP_PROP_FRAME_WIDTH, frame_width)
        cap.set(cv2.CAP_PROP_FRAME_HEIGHT, frame_height)
        model = YOLO("yolov8l.pt")
        box_annotator = sv.BoxAnnotator(
            thickness=2,
            text_thickness=2,
            text_scale=1
        )
        zone_polygon = (ZONE_POLYGON * np.array(args.webcam_resolution)).astype(int)
        zone = sv.PolygonZone(polygon=zone_polygon, frame_resolution_wh=tuple(args.webcam_resolution))
        zone_annotator = sv.PolygonZoneAnnotator(
            zone=zone,
            color=sv.Color.red(),
            thickness=2,
            text_thickness=4,
            text_scale=2
        )
        # Initialize the flags to False
        min_message_sent = False
        max_message_sent = False
        while True:
            ret, frame = cap.read()
            result = model(frame, agnostic_nms=True)[0]
            detections = sv.Detections.from_yolov8(result)
            detections = detections[detections.class_id == 0]
            labels = [
                f"{model.model.names[class_id]} {confidence:0.2f}"
                for _, confidence, class_id, _
                in detections
            ]
            frame = box_annotator.annotate(
                scene=frame,
                detections=detections,
                labels=labels
            )
            zone.trigger(detections=detections)
            frame = zone_annotator.annotate(scene=frame)
            person_count = len(detections)
            #person_count = 0
            for detection in detections:
                x, y, w, h = detection

                if h is not None:
                    center_x = x + w / 2
                    center_y = y + h / 2

                 
                    if zone.contains_point([center_x, center_y]):
                        person_count += 1

            # Check the flags before sending messages
            if person_count <= MIN_PERSON_COUNT and not min_message_sent:
                min_message_sent = True  # Update the flag to True
                bot.sendMessage(receiver_id, "Minimum number of people reached.")

            if person_count >= MAX_PERSON_COUNT and not max_message_sent:
                #bot.sendMessage(receiver_id, "Maximum kişi sayısına ulaşıldı")
                max_message_sent = True  # Update the flag to True
                bot.sendMessage(receiver_id, "Maximum number of people reached.")

            # Reset the flags when the person count changes
            if person_count > MIN_PERSON_COUNT and min_message_sent:
                min_message_sent = False  # Reset the flag to False

            if person_count < MAX_PERSON_COUNT and max_message_sent:
                max_message_sent = False  # Reset the flag to False

            ret, buffer = cv2.imencode('.jpg', frame)
            frame_bytes = buffer.tobytes()

......

Thanks for your attention :)

[hardikdava]

@yavuzhantemiz Thank you for using supervision and taking interest in improving supervision. Your suggestion is too specific. Just keep in a mind that supervision only plays intermediate role to build your computer vision application. Currently, we are not working on this API but in future we are going to improve it. But we will definitely consider your suggestion when we start working on it. Thank you again for taking interest in supervision.