hanium_detect.py

# 9월 6일 그 전에 구현된 경보, 메시지 기능 UI 연동 구현 위해서 다시 추가, UI 관련으로 detect 파라미터 추가

import datetime
import argparse
import os
import time
from pathlib import Path
from threading import Thread, Timer

from numpy import True_
from siren import call_siren
import cv2
import torch
import torch.backends.cudnn as cudnn
import os
import winsound
import threading
import usb.core
import usb.util
import pymysql
from kakao import send_message
from learning import deepcall
from models.experimental import attempt_load
# from siren import call_siren
from utils.datasets import LoadStreams, LoadImages
from utils.general import check_img_size, check_requirements, check_imshow, non_max_suppression, apply_classifier, \
    scale_coords, xyxy2xywh, strip_optimizer, set_logging, increment_path, save_one_box
from utils.plots import colors, plot_one_box
from utils.torch_utils import select_device, load_classifier, time_synchronized


def check(xyxy, xyxytemp):
    if xyxy[0] - 10 > xyxytemp[2]:
        return False
    elif xyxy[2] + 10 < xyxytemp[0]:
        return False
    elif xyxy[3] + 10 < xyxytemp[1]:
        return False
    elif xyxy[1] - 10 > xyxytemp[3]:
        return False
    else:
        return True


def sani_pos_check(xyxy, sani_pos):
    if xyxy[0] > sani_pos[2]:
        return False
    elif xyxy[2] < sani_pos[0]:
        return False
    elif xyxy[3] < sani_pos[1]:
        return False
    elif xyxy[1] > sani_pos[3]:
        return False
    else:
        return True


def temp_pos_check(xyxy, temp_pos):
    if xyxy[0] > temp_pos[2]:
        return False
    elif xyxy[2] < temp_pos[0]:
        return False
    elif xyxy[3] < temp_pos[1]:
        return False
    elif xyxy[1] > temp_pos[3]:
        return False
    else:
        return True


def qr_pos_check(xyxy, qr_pos):
    if xyxy[0] > qr_pos[2]:
        return False
    elif xyxy[2] < qr_pos[0]:
        return False
    elif xyxy[3] < qr_pos[1]:
        return False
    elif xyxy[1] > qr_pos[3]:
        return False
    else:
        return True


def check_Cross(x, comp):
    if comp + 30 > x > comp - 30:
        return True
    else:
        return False


# --------------텍스트박스 위치--------------------#
center = [960, 100, 960, 100]
tmp = [100, 100, 100, 100]
tmp_mask = [1600, 50, 1600, 50]
tmp_sani = [1600, 100, 1600, 100]
tmp_temp = [1600, 150, 1600, 150]
tmp_qrcd = [1600, 200, 1600, 200]
exit = [1600, 250, 1600, 250]
siren = [200, 700, 200, 700]
# ---------------텍스트박스 위치-------------------#
deepcall_check = [0, 0, 0, 0]  # 객체 검출이 특정횟수 이상 연속으로 검출되어야 사용했다고 판정하기 위해 사용
detected_sani_count = [0]  # sani의 검출 횟수를 담는 변수
detected_temp_count = [0]  # temp의 검출 횟수를 담는 변수
detected_qr_count = [0]  # qr의 검출 횟수를 담는 변수
detected_mask_count = [0]  # mask의 검출 횟수를 담는 변수
log_data = []

juso_db = pymysql.connect(
    user='root',
    passwd='root',
    host='127.0.0.1',
    db='corona',
    charset='utf8'
)


@torch.no_grad()
def detect(weights='./8192_weights',  # model.pt path(s)
           source='data/images',  # file/dir/URL/glob, 0 for webcam
           w_width=1280,
           w_height=720,
           imgsz=640,  # inference size (pixels)
           conf_thres=0.25,  # confidence threshold
           iou_thres=0.45,  # NMS IOU threshold
           max_det=1000,  # maximum detections per image
           device='',  # cuda device, i.e. 0 or 0,1,2,3 or cpu
           view_img=False,  # show results
           save_txt=False,  # save results to *.txt
           save_conf=False,  # save confidences in --save-txt labels
           save_crop=False,  # save cropped prediction boxes
           nosave=False,  # do not save images/videos
           classes=None,  # filter by class: --class 0, or --class 0 2 3
           agnostic_nms=False,  # class-agnostic NMS
           augment=False,  # augmented inference
           update=False,  # update all models
           project='runs/detect',  # save results to project/name
           name='exp',  # save results to project/name
           exist_ok=False,  # existing project/name ok, do not increment
           line_thickness=3,  # bounding box thickness (pixels)
           hide_labels=False,  # hide labels
           hide_conf=False,  # hide confidences
           half=False,  # use FP16 half-precision inference
           mod=0,
           set_alarm=0
           ):
    sani_pos = [1550, 380, 1650, 480]
    temp_pos = [950, 270, 1050, 370]
    qr_pos = [420, 260, 520, 360]

    # 알람 모두 사용
    alarm_light = True
    alarm_siren = True
    alarm_msg = True

    if set_alarm == 1:  # siren 제외
        alarm_siren = False
    elif set_alarm == 2:  # light 제외
        alarm_light = False
    elif set_alarm == 3:  # msg 제외
        alarm_msg = False
    elif set_alarm == 4:  # siren만
        alarm_light = False
        alarm_msg = False
    elif set_alarm == 5:  # light만
        alarm_siren = False
        alarm_msg = False
    elif set_alarm == 6:  # msg만
        alarm_siren = False
        alarm_light = False
    elif set_alarm == 7:
        alarm_msg = False
        alarm_light = False
        alarm_siren = False

    mode_check = []
    if mod == 0:
        mode_check.append(0.0)
        mode_check.append(1.0)
        mode_check.append(2.0)
        mode_check.append(3.0)
        mode_check.append(4.0)
        mode_check.append(5.0)
    elif mod == 1:
        mode_check.append(0.0)
        mode_check.append(1.0)
        mode_check.append(3.0)
        mode_check.append(4.0)
        mode_check.append(5.0)
    elif mod == 2:
        mode_check.append(0.0)
        mode_check.append(1.0)
        mode_check.append(2.0)
        mode_check.append(4.0)
        mode_check.append(5.0)
    elif mod == 3:
        mode_check.append(0.0)
        mode_check.append(1.0)
        mode_check.append(2.0)
        mode_check.append(3.0)
        mode_check.append(5.0)
    elif mod == 4:
        mode_check.append(0.0)
        mode_check.append(1.0)
        mode_check.append(2.0)
        mode_check.append(5.0)
    elif mod == 5:
        mode_check.append(0.0)
        mode_check.append(1.0)
        mode_check.append(3.0)
        mode_check.append(5.0)
    elif mod == 6:
        mode_check.append(0.0)
        mode_check.append(1.0)
        mode_check.append(4.0)
        mode_check.append(5.0)

    now = datetime.datetime.now()
    print(" '%s' ", now)

    cursor = juso_db.cursor(pymysql.cursors.DictCursor)

    db_check = "SHOW TABLES LIKE 'log';"
    cursor.execute(db_check)
    db_check_result = cursor.fetchall()
    if len(db_check_result) == 0:
        cursor.execute("""CREATE TABLE log(
                id INT(255) NOT NULL AUTO_INCREMENT PRIMARY KEY, 
                time DATETIME, 
                check_act VARCHAR(255),
                ab_path VARCHAR(255),
                file_name VARCHAR(255)
                );""")
        cursor.fetchall()

    check_sani = False  # BBOX 겹침이 발생했을 때, sani의 bbox를 custom-Layer로 보내서, 손소독제를 짜는 상황의 sani인지 체크하는 변수
    check_temp = False  # BBOX 겹침이 발생했을 때, temp의 bbox를 custom-Layer로 보내서, 열을 재고있는 temp인지 체크하는 변수
    check_qrcd = False  # BBOX 겹침이 발생했을 때, qrcd의 bbox를 custom-Layer로 보내서, qr검사를 하고 있는 qr인지 체크하는 변수
    start_x = 0  # 사람이 입장하고, 마스크착용 검사를 하는 x좌표
    sani_x_start = 0
    temp_x_start = 0
    qrcd_x_start = 0
    sani_x_end = 0  # sani_check의 값에 따라 경보를 울릴것인지, 말 것인지 결정하는 위치를 담음
    temp_x_end = 0  # temp_check의 값에 따라 경보를 울릴것인지, 말 것인지 결정하는 위치를 담음
    qrcd_x_end = 0  # qrcd_check의 값에 따라 경보를 울릴것인지, 말 것인지 결정하는 위치를 담음
    key = True  # False = 설정모드, True = 검출모드
    init_check = [0, 0, 0]  # 3객체가 적당한 위치에 배치되었는지 확인하는 용도, [1,1,1]이 저장된다면 key를 true로 바꾸고 검출모드 시작
    init_check = [0, 0, 0]  # 3객체가 적당한 위치에 배치되었는지 확인하는 용도, [1,1,1]이 저장된다면 key를 true로 바꾸고 검출모드 시작
    sani_lock = [False, False]  # sani의 검출이 sani_x 주변에서 딱 1번만 실행하도록 하는 용도
    temp_lock = [False, False]  # temp의 검출이 temp_x 주변에서 딱 1번만 실행하도록 하는 용도
    qr_lock = [False, False]  # qr의 검출이 qr 주변에서 딱 1번만 실행하도록 하는 용도
    checking = 0
    save_img = not nosave and not source.endswith('.txt')  # save inference images
    webcam = source.isnumeric() or source.endswith('.txt') or source.lower().startswith(
        ('rtsp://', 'rtmp://', 'http://', 'https://'))
    # Directories
    save_dir = increment_path(Path(project) / name, exist_ok=exist_ok)  # increment run
    (save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True)  # make dir

    def sani_lock_free():
        sani_lock[0] = False
        sani_lock[1] = False

    def temp_lock_free():
        temp_lock[0] = False
        temp_lock[1] = False

    def qr_lock_free():
        qr_lock[0] = False
        qr_lock[1] = False

    # Initialize
    set_logging()
    device = select_device(device)
    half &= device.type != 'cpu'  # half precision only supported on CUDA

    # Load model
    model = attempt_load(weights, map_location=device)  # load FP32 model
    stride = int(model.stride.max())  # model stride
    imgsz = check_img_size(imgsz, s=stride)  # check image size
    names = model.module.names if hasattr(model, 'module') else model.names  # get class names
    if half:
        model.half()  # to FP16

    # Second-stage classifier
    classify = False
    if classify:
        modelc = load_classifier(name='resnet101', n=2)  # initialize
        modelc.load_state_dict(torch.load('weights/resnet101.pt', map_location=device)['model']).to(device).eval()

    # Set Dataloader
    vid_path, vid_writer = None, None
    if webcam:
        view_img = check_imshow()
        cudnn.benchmark = True  # set True to speed up constant image size inference
        dataset = LoadStreams(source, img_size=imgsz, stride=stride, w_width=w_width, w_height=w_height)
    else:
        view_img = check_imshow()
        dataset = LoadImages(source, img_size=imgsz, stride=stride)

    for dirpath, dirnames, filenames in os.walk('runs/detect/sanitizer'):
        # Remove regular files, ignore directories
        for filename in filenames:
            os.unlink(os.path.join(dirpath, filename))
    for dirpath, dirnames, filenames in os.walk('runs/detect/temperature'):
        # Remove regular files, ignore directories
        for filename in filenames:
            os.unlink(os.path.join(dirpath, filename))
    for dirpath, dirnames, filenames in os.walk('runs/detect/qr'):
        # Remove regular files, ignore directories
        for filename in filenames:
            os.unlink(os.path.join(dirpath, filename))
    # Run inference
    if device.type != 'cpu':
        model(torch.zeros(1, 3, imgsz, imgsz).to(device).type_as(next(model.parameters())))  # run once
    t0 = time.time()
    for path, img, im0s, vid_cap in dataset:
        img = torch.from_numpy(img).to(device)
        img = img.half() if half else img.float()  # uint8 to fp16/32
        img /= 255.0  # 0 - 255 to 0.0 - 1.0
        if img.ndimension() == 3:
            img = img.unsqueeze(0)

        # Inference
        t1 = time_synchronized()
        pred = model(img, augment=augment)[0]

        # Apply NMS
        pred = non_max_suppression(pred, conf_thres, iou_thres, classes, agnostic_nms, max_det=max_det)
        t2 = time_synchronized()

        # Apply Classifier
        if classify:
            pred = apply_classifier(pred, modelc, img, im0s)

        # Process detections
        for i, det in enumerate(pred):  # detections per image
            if webcam:  # batch_size >= 1
                p, s, im0, frame = path[i], f'{i}: ', im0s[i].copy(), dataset.count
            else:
                p, s, im0, frame = path, '', im0s.copy(), getattr(dataset, 'frame', 0)
            p = Path(p)  # to Path
            save_path = str(save_dir / p.name)  # img.jpg
            txt_path = str(save_dir / 'labels' / p.stem) + ('' if dataset.mode == 'image' else f'_{frame}')  # img.txt
            s += '%gx%g ' % img.shape[2:]  # print string
            gn = torch.tensor(im0.shape)[[1, 0, 1, 0]]  # normalization gain whwh
            imc = im0.copy() if save_crop else im0  # for save_crop
            if len(det):
                # Rescale boxes from img_size to im0 size
                det[:, :4] = scale_coords(img.shape[2:], det[:, :4], im0.shape).round()

                # Print results
                for c in det[:, -1].unique():
                    n = (det[:, -1] == c).sum()  # detections per class
                    s += f"{n} {names[int(c)]}{'s' * (n > 1)}, "  # add to string
                # Write results
                if key == False:
                    if mod == 0:
                        if init_check != [1, 1, 1]:
                            init_check = [0, 0, 0]
                            checking = 0
                        else:
                            init_check = [0, 0, 0]
                            checking = checking + 1
                            cv2.putText(im0, "%d" % (10 - checking / 20), (700, 700), cv2.FONT_ITALIC, 30,
                                        (0, 140, 255), 50)
                    elif mod == 1:
                        if init_check != [0, 1, 1]:
                            init_check = [0, 0, 0]
                        else:
                            init_check = [0, 0, 0]
                            checking = checking + 1
                            cv2.putText(im0, "%d" % (10 - checking / 20), (500, 500), cv2.FONT_ITALIC, 30,
                                        (255, 140, 0), 5)
                    elif mod == 2:
                        if init_check != [1, 0, 1]:
                            init_check = [0, 0, 0]
                            checking = 0
                        else:
                            init_check = [0, 0, 0]
                            checking = checking + 1
                            cv2.putText(im0, "%d" % (10 - checking / 20), (500, 500), cv2.FONT_ITALIC, 30,
                                        (255, 140, 0), 5)
                    elif mod == 3:
                        if init_check != [1, 1, 0]:
                            init_check = [0, 0, 0]
                            checking = 0
                        else:
                            init_check = [0, 0, 0]
                            checking = checking + 1
                            cv2.putText(im0, "%d" % (10 - checking / 20), (
                            int(im0.shape[1]/ 2),
                            int(im0.shape[0] / 2)), cv2.FONT_ITALIC, 30, (255, 255, 255), 5)
                    elif mod == 4:
                        if init_check != [1, 0, 0]:
                            init_check = [0, 0, 0]
                            checking = 0
                        else:
                            init_check = [0, 0, 0]
                            checking = checking + 1
                            cv2.putText(im0, "%d" % (10 - checking / 20), (
                            int(im0.shape[1]/ 2),
                            int(im0.shape[0] / 2)), cv2.FONT_ITALIC, 30, (255, 255, 255), 5)
                    elif mod == 5:
                        if init_check != [0, 1, 0]:
                            init_check = [0, 0, 0]
                            checking = 0
                        else:
                            init_check = [0, 0, 0]
                            checking = checking + 1
                            cv2.putText(im0, "%d" % (10 - checking / 20), (
                            int(im0.shape[1]/ 2),
                            int(im0.shape[0] / 2)), cv2.FONT_ITALIC, 30, (255, 255, 255), 5)
                    elif mod == 6:
                        if init_check != [0, 0, 1]:
                            init_check = [0, 0, 0]
                            checking = 0
                        else:
                            init_check = [0, 0, 0]
                            checking = checking + 1
                            cv2.putText(im0, "%d" % (10 - checking / 20), (
                            int(im0.shape[1]/ 2),
                            int(im0.shape[0] / 2)), cv2.FONT_ITALIC, 30, (255, 255, 255), 5)
                if (checking / 20) == 10:
                    key = True
                for *xyxy, conf, cls in reversed(
                        det):  # reversed(det) = 1box, cls 0.0 = head , 1.0 = hands, 2.0 = sanitizer, 3.0 = temperature, 4.0 = qrcd
                    if cls.item() in mode_check:
                        if cls.item() == 2.0 and sani_pos_check(xyxy, sani_pos):
                            init_check[0] = 1                    
                        if cls.item() == 3.0 and temp_pos_check(xyxy, temp_pos):
                            init_check[1] = 1
                        if cls.item() == 4.0 and qr_pos_check(xyxy, qr_pos):
                            init_check[2] = 1
                        if save_txt:  # Write to file
                            xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist()  # normalized xywh
                            line = (cls, *xywh, conf) if save_conf else (cls, *xywh)  # label format
                            with open(txt_path + '.txt', 'a') as f:
                                f.write(('%g ' * len(line)).rstrip() % line + '\n')
                        if save_img or save_crop or view_img:  # Add bbox to image
                            c = int(cls)  # integer class
                            label = None if hide_labels else (names[c] if hide_conf else f'{names[c]} {conf:.2f}')
                            plot_one_box(xyxy, im0, label=label, color=colors(c, True), line_thickness=line_thickness)
                            if save_crop:
                                save_one_box(xyxy, imc, file=save_dir / 'crops' / names[c] / f'{p.stem}.jpg', BGR=True)
                        # 여기부터 우리가 추가한 코드이고, 3객체를 올바른 위치에 배치시키면 key를 True로 바꾸면서 검출모드가 시작된다.
                        if key:
                            plot_one_box(center, im0, label="ON DETECTING", color=colors(int(0), True),
                                         line_thickness=line_thickness)
                            if cls.item() == 2.0:
                                sani_x_start = int(xyxy[0] + 100)
                                sani_x_end = int(xyxy[0] - 100)
                            if cls.item() == 3.0:
                                temp_x_start = int(xyxy[0] + 100)
                                temp_x_end = int(xyxy[0] - 100)
                            if cls.item() == 4.0:
                                qrcd_x_start = int(xyxy[0] + 100)
                                qrcd_x_end = int(xyxy[0] - 100)
                            start_x = int(im0.shape[1]- 100) 

                            cv2.line(im0, (start_x, 0), (start_x, 1000), (255, 255, 255), 1)
                            cv2.line(im0, (sani_x_end, 0), (sani_x_end, 1000), (0, 0, 255), 1)  # red   
                            cv2.line(im0, (temp_x_end, 0), (temp_x_end, 1000), (0, 255, 0),
                                     1)  # green
                            cv2.line(im0, (qrcd_x_end, 0), (qrcd_x_end, 1000), (255, 0, 0),
                                     1)  # blue
                            if (sani_lock[0] == False):
                                cv2.line(im0, (sani_x_start, 0), (sani_x_start, 1000), (0, 0, 255), 1)  # red

                            if (temp_lock[0] == False):
                                cv2.line(im0, (temp_x_start, 0), (temp_x_start, 1000), (0, 255, 0), 1)  # green 

                            if (qr_lock[0] == False):
                                cv2.line(im0, (qrcd_x_start, 0), (qrcd_x_start, 1000), (255, 0, 0), 1)  # blue

                            if cls.item() == 0.0 or cls.item() == 1.0:
                                # check mask
                                if cls.item() == 1.0:
                                    if start_x > xyxy[0] > start_x - 50 and 5.0 in mode_check:
                                        x1 = xyxy[0]
                                        x2 = xyxy[2]
                                        y1 = xyxy[1]
                                        y2 = xyxy[3]
                                        img_trim = im0s[0][int(y1):int(y2), int(x1):int(x2)]
                                        cv2.imwrite("./tmp/img/1/out.jpg", img_trim)
                                        #if deepcall() == 6:
                                        #    plot_one_box(tmp, im0, label="check = mask", color=colors(int(cls), True),
                                        #                 line_thickness=line_thickness)
                                        #else:
                                        plot_one_box(tmp, im0, label="check = not mask",
                                                        color=colors(int(cls), True),
                                                         line_thickness=line_thickness)
                                        deepcall_check[3] = deepcall_check[3] + 1
                                        if deepcall_check[3] == 5:
                                            detected_mask_count[0] = detected_mask_count[0] + 1
                                            plot_one_box(siren, im0, label="Not Mask!!!",
                                                         color=colors(int(200), True),
                                                         line_thickness=line_thickness)
                                            file_name = "mask" + str(detected_mask_count[0])
                                            repath = "./detected_image/mask/" + file_name + ".jpg"
                                            cv2.imwrite(repath, im0)
                                            ab_path = Path(repath).absolute()
                                            now = datetime.datetime.now()
                                            now = now.strftime('%Y-%m-%d %H:%M:%S')
                                            log_data.append(
                                                {
                                                    "time": now,
                                                    "act": "마스크 미착용",
                                                }
                                            )
                                            cursor.execute("""INSERT INTO log (time, check_act, ab_path, file_name) VALUES (%s,'마스크',%s,%s)""", (now, ab_path, file_name))
                                            if alarm_msg:
                                                send_message(now, "7호관 뒷문 검역소", "마스크")
                                            cursor.fetchall()
                                            juso_db.commit()
                                            if alarm_light or alarm_siren:
                                                th1 = Thread(target=call_siren)
                                                th1.start()                                                   
                                            deepcall_check[3] = 0
                                        

                                for *xyxytmp, clstmp in reversed(det):
                                    if clstmp.item() in mode_check:
                                        if clstmp.item() != 0.0 and clstmp.item() != 1.0 and check(xyxy,
                                                                                                   xyxytmp):  # cls(손, 얼굴) clstmp(손소독,qr,온도계)
                                            if cls.item() == 0.0 and clstmp.item() == 2.0:
                                                x1 = xyxytmp[0]
                                                x2 = xyxytmp[2]
                                                y1 = xyxytmp[1]
                                                y2 = xyxytmp[3]
                                                ry = (y2 - y1) / 8
                                                y1 = int(y1 - ry)
                                                y2 = int(y2)
                                                x1 = int(x1)
                                                x2 = int(x2)
                                                img_trim = im0s[0][y1:y2, x1:x2]
                                                cv2.imwrite("./tmp/img/1/out.jpg", img_trim)
                                                if len(os.listdir("./tmp/img/1")) != 0:
                                                    if deepcall() == 0:
                                                        plot_one_box(tmp, im0, label="check = sanitizer",
                                                                     color=colors(int(cls), True),
                                                                     line_thickness=line_thickness)
                                                        deepcall_check[0] = deepcall_check[0] + 1
                                                        if deepcall_check[0] == 5:
                                                            check_sani = True
                                                            deepcall_check[0] = 0
                                            elif cls.item() == 1.0 and clstmp.item() == 3.0:
                                                x1 = xyxytmp[0]
                                                x2 = xyxytmp[2]
                                                y1 = xyxytmp[1]
                                                y2 = xyxytmp[3]
                                                rx = (x2 - x1) / 2
                                                y1 = int(y1)
                                                y2 = int(y2)
                                                x1 = int(x1 - rx)
                                                x2 = int(x2 + rx)
                                                img_trim = im0s[0][y1:y2, x1:x2]
                                                cv2.imwrite("./tmp/img/1/out.jpg", img_trim)
                                                if len(os.listdir("./tmp/img/1")) != 0:
                                                    if deepcall() == 2:
                                                        plot_one_box(tmp, im0, label="check = temperatrue",
                                                                     color=colors(int(cls), True),
                                                                     line_thickness=line_thickness)
                                                        deepcall_check[1] = deepcall_check[1] + 1
                                                        if deepcall_check[1] == 5:
                                                            check_temp = True
                                                            deepcall_check[1] = 0
                                            elif cls.item() == 0.0 and clstmp.item() == 4.0:
                                                x1 = xyxytmp[0]
                                                x2 = xyxytmp[2]
                                                y1 = xyxytmp[1]
                                                y2 = xyxytmp[3]
                                                rx = (x2 - x1) / 2
                                                y1 = int(y1)
                                                y2 = int(y2)
                                                x1 = int(x1 - rx)
                                                x2 = int(x2 + rx)
                                                img_trim = im0s[0][y1:y2, x1:x2]
                                                cv2.imwrite("./tmp/img/1/out.jpg", img_trim)
                                                if len(os.listdir("./tmp/img/1")) != 0:
                                                    if deepcall() == 4:
                                                        plot_one_box(tmp, im0, label="check = qrcode",
                                                                     color=colors(int(cls), True),
                                                                     line_thickness=line_thickness)
                                                        deepcall_check[2] = deepcall_check[2] + 1
                                                        if deepcall_check[2] == 5:
                                                            check_qrcd = True
                                                            deepcall_check[2] = 0

                            if cls.item() == 1.0:
                                if check_Cross(xyxy[0], sani_x_start) and sani_lock[0] == False:
                                    check_sani = False
                                    sani_lock[0] = True
                                    # threading.Timer(20, sani_lock_free).start()
                                if check_Cross(xyxy[0], temp_x_start) and temp_lock[0] == False:
                                    check_temp = False
                                    temp_lock[0] = True
                                    # threading.Timer(10, temp_lock_free).start()
                                if check_Cross(xyxy[0], qrcd_x_start) and qr_lock[0] == False:
                                    check_qrcd = False
                                    qr_lock[0] = True
                                    # threading.Timer(10, qr_lock_free).start()

                                if check_Cross(xyxy[2], sani_x_end) and 2.0 in mode_check:
                                    sani_lock[1] = False
                                if check_Cross(xyxy[2], temp_x_end) and 3.0 in mode_check:
                                    temp_lock[1] = False
                                if check_Cross(xyxy[2], qrcd_x_end) and 4.0 in mode_check:
                                    qr_lock[1] = False
                                if check_Cross(xyxy[0], sani_x_end) and 2.0 in mode_check:
                                    sani_lock[0] = False
                                    if check_sani == False and sani_lock[1] == False:
                                        plot_one_box(siren, im0, label="Not Sani!!!", color=colors(int(200), True),
                                                     line_thickness=line_thickness)
                                        detected_sani_count[0] = detected_sani_count[0] + 1
                                        file_name = "sani" + str(detected_sani_count[0])
                                        repath = "./detected_image/sani/" + file_name + ".jpg"
                                        cv2.imwrite(repath, im0)
                                        ab_path = Path(repath).absolute()
                                        now = datetime.datetime.now()
                                        now = now.strftime('%Y-%m-%d %H:%M:%S')
                                        log_data.append(
                                            {
                                                "time": now,
                                                "act": "손소독제 미사용",
                                            }
                                        )
                                        cursor.execute("""INSERT INTO log (time, check_act, ab_path, file_name) VALUES (%s,'손소독',%s,%s)""", (now, ab_path, file_name))
                                        if alarm_msg:
                                           send_message(now, "7호관 뒷문 검역소", "손소독제")
                                        cursor.fetchall()
                                        juso_db.commit()
                                        if alarm_light or alarm_siren:
                                            th1 = Thread(target=call_siren)
                                            th1.start()
                                        sani_lock[1] = True
                                        # threading.Timer(10, sani_lock_free).start()

                                if check_Cross(xyxy[0], temp_x_end) and 3.0 in mode_check:
                                    temp_lock[0] = False
                                    if check_temp == False and temp_lock[1] == False:
                                        plot_one_box(siren, im0, label="Not temp!!!", color=colors(int(200), True),
                                                     line_thickness=line_thickness)
                                        detected_temp_count[0] = detected_temp_count[0] + 1
                                        file_name = "temp" + str(detected_temp_count[0])
                                        repath = "./detected_image/temp/" + file_name + ".jpg"
                                        cv2.imwrite(repath, im0)
                                        ab_path = Path(repath).absolute()
                                        now = datetime.datetime.now()
                                        now = now.strftime('%Y-%m-%d %H:%M:%S')
                                        log_data.append(
                                            {
                                                "time": now,
                                                "act": "온도계 미사용",
                                            }
                                        )
                                        cursor.execute("""INSERT INTO log (time, check_act, ab_path, file_name) VALUES (%s,'온도계',%s,%s)""", (now, ab_path, file_name))
                                        if alarm_msg:
                                            send_message(now, "7호관 뒷문 검역소", "체온검사")
                                        cursor.fetchall()
                                        juso_db.commit()
                                        if alarm_light or alarm_siren:
                                            th1 = Thread(target=call_siren)
                                            th1.start()
                                        temp_lock[1] = True
                                        # threading.Timer(10, temp_lock_free).start()

                                if check_Cross(xyxy[0], qrcd_x_end) and 4.0 in mode_check:
                                    qr_lock[0] = False
                                    if check_qrcd == False and qr_lock[1] == False:
                                        plot_one_box(siren, im0, label="Not qrcd!!!", color=colors(int(200), True),
                                                     line_thickness=line_thickness)
                                        detected_qr_count[0] = detected_qr_count[0] + 1
                                        file_name = "qrcd" + str(detected_qr_count[0])
                                        repath = "./detected_image/qr/" + file_name + ".jpg"
                                        cv2.imwrite(repath, im0)
                                        ab_path = Path(repath).absolute()
                                        now = datetime.datetime.now()
                                        now = now.strftime('%Y-%m-%d %H:%M:%S')
                                        log_data.append(
                                            {
                                                "time": now,
                                                "act": "QR 코드 체크 미시행",
                                            }
                                        )
                                        cursor.execute("""INSERT INTO log (time, check_act, ab_path, file_name) VALUES (%s,'QR',%s,%s)""", (now, ab_path, file_name))
                                        if alarm_msg:
                                           send_message(now, "7호관 뒷문 검역소", "qr")
                                        cursor.fetchall()
                                        juso_db.commit()
                                        if alarm_light or alarm_siren:
                                            th1 = Thread(target=call_siren)
                                            th1.start()
                                        qr_lock[1] = True
                                        # threading.Timer(10, qr_lock_free).start()
                    if key:  # key=true로 설정된 이후에 보여지는 것들입니다.
                        if 5.0 in mode_check:
                            plot_one_box(tmp_mask, im0, label="mask_detect = %d" % detected_mask_count[0],
                                         color=colors(int(200), True),
                                         line_thickness=line_thickness)
                        if 2.0 in mode_check:
                            plot_one_box(tmp_sani, im0, label="sani_detect = %d" % detected_sani_count[0],
                                         color=colors(int(200), True),
                                         line_thickness=line_thickness)
                        if 3.0 in mode_check:
                            plot_one_box(tmp_temp, im0, label="temp_detect = %d" % detected_temp_count[0],
                                         color=colors(int(200), True),
                                         line_thickness=line_thickness)
                        if 4.0 in mode_check:
                            plot_one_box(tmp_qrcd, im0, label="qrcd_detect = %d" % detected_qr_count[0],
                                         color=colors(int(200), True),
                                         line_thickness=line_thickness)
                    # 여기까지가 우리가 수정한 부분입니다.

            # Print time (inference + NMS)
            print(f'{s}Done. ({t2 - t1:.3f}s)')
            plot_one_box(exit, im0, label="EXIT : Q", color=colors(int(200), True),
                         line_thickness=line_thickness)
            if key == False:
                plot_one_box(center, im0, label="SETTING.....", color=colors(int(0), True),
                             line_thickness=line_thickness)
                if 2.0 in mode_check:
                    plot_one_box(sani_pos, im0, label="Place Sani", color=colors(0, True),
                                 line_thickness=line_thickness)
                if 3.0 in mode_check:
                    plot_one_box(temp_pos, im0, label="Place Temp", color=colors(127, True),
                                 line_thickness=line_thickness)
                if 4.0 in mode_check:
                    plot_one_box(qr_pos, im0, label="Place QR", color=colors(255, True),
                                 line_thickness=line_thickness)

                    # Stream results
            if view_img:
                cv2.imshow(str(p), im0)
                cv2.waitKey(1)  # 1 millisecond

            # Save results (image with detections)
            if save_img:
                if dataset.mode == 'image':
                    cv2.imwrite(save_path, im0)
                else:  # 'video' or 'stream'
                    if vid_path != save_path:  # new video
                        vid_path = save_path
                        if isinstance(vid_writer, cv2.VideoWriter):
                            vid_writer.release()  # release previous video writer
                        if vid_cap:  # video
                            fps = vid_cap.get(cv2.CAP_PROP_FPS)
                            w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH))
                            h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
                        else:  # stream
                            fps, w, h = 30, im0.shape[1], im0.shape[0]
                            save_path += '.mp4'
                        vid_writer = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h))
                    vid_writer.write(im0)

    if save_txt or save_img:
        s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else ''
        print(f"Results saved to {save_dir}{s}")

    if update:
        strip_optimizer(weights)  # update model (to fix SourceChangeWarning)

    print(f'Done. ({time.time() - t0:.3f}s)')


if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--weights', nargs='+', type=str, default='yolov5s.pt', help='model.pt path(s)')
    parser.add_argument('--source', type=str, default='data/images', help='file/dir/URL/glob, 0 for webcam')
    parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='inference size (pixels)')
    parser.add_argument('--conf-thres', type=float, default=0.25, help='confidence threshold')
    parser.add_argument('--iou-thres', type=float, default=0.45, help='NMS IoU threshold')
    parser.add_argument('--max-det', type=int, default=1000, help='maximum detections per image')
    parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
    parser.add_argument('--view-img', action='store_true', help='show results')
    parser.add_argument('--save-txt', action='store_true', help='save results to *.txt')
    parser.add_argument('--save-conf', action='store_true', help='save confidences in --save-txt labels')
    parser.add_argument('--save-crop', action='store_true', help='save cropped prediction boxes')
    parser.add_argument('--nosave', action='store_true', help='do not save images/videos')
    parser.add_argument('--classes', nargs='+', type=int, help='filter by class: --class 0, or --class 0 2 3')
    parser.add_argument('--agnostic-nms', action='store_true', help='class-agnostic NMS')
    parser.add_argument('--augment', action='store_true', help='augmented inference')
    parser.add_argument('--update', action='store_true', help='update all models')
    parser.add_argument('--project', default='runs/detect', help='save results to project/name')
    parser.add_argument('--name', default='exp', help='save results to project/name')
    parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment')
    parser.add_argument('--line-thickness', default=3, type=int, help='bounding box thickness (pixels)')
    parser.add_argument('--hide-labels', default=False, action='store_true', help='hide labels')
    parser.add_argument('--hide-conf', default=False, action='store_true', help='hide confidences')
    parser.add_argument('--half', action='store_true', help='use FP16 half-precision inference')
    opt = parser.parse_args()
    check_requirements(exclude=('tensorboard', 'thop'))
    detect(**vars(opt))