#进入线程先把标志位置为True
# from System_I import Detection
# import os,torch
# os.environ['CUDA_VISIBLE_DEVICES'] = '1'
# devices = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
# root = "/home/wanghc/output"
import  os,torch, threading,argparse
class global_para():
    parser1 = argparse.ArgumentParser(description="GYD")
    parser1.add_argument("video_num", default='0001', type=str,
                        help='Video name.')
    parser1.add_argument("GPU", default='5', type=str,
                        help='GPU.')
    parser1.add_argument("thresh", default=0.6, type=float,
                        help='GPU.')
    parser1.add_argument("red", default=0, type=int,
                        help='red.')

    args = parser1.parse_args()
    print(args)
    red_flag = args.red
    detect_thresh = args.thresh
    is_ex_bird = False  # 目标检测标志位，每个一秒检测一次目标检测，如果有鸟，则置为True，为True的时候说明， 处理30s的线程还在继续，直到标志位变为了False再开启新的线程
    Density_thread_flag = False  # 密度估计线程标志位，为True说明开启了密度估计线程，处理完750帧为止，标志位标为False，
    Detection_thread_flag = False  # 目标检测和细粒度识别线程标志位，为True说明开启了密度估计线程，处理完750帧为止，标志位标为False，
    Vedio_thread_flag = False  # 存储视频线程标志位，存储30s
    SaveImage_thread_flag = False  # 保存原图线程标志位，存储30张原图
    ALL_thread_flag = False  # 所有线程的标志位，单独开设一个线程，监控所有线程是否都结束，如果都结束，则置为False，后续添加线程管理，可以放在这里面

    DebugPrint_flag = True  # 打印标志位

    destory_image = False
    Camera_device = args.video_num  # 摄像头编号，0001
    image_time = None  # 当前处理的30视频的编号
    ImgDealQueue = []  # 750帧视频，缓存空间
    root = "/data/Hatcher/output2"
    #os.environ['CUDA_VISIBLE_DEVICES'] = '0'
    os.environ['CUDA_VISIBLE_DEVICES'] = args.GPU
    devices = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
    #event = threading.Event()
    eventDensity = threading.Event()
    eventDetection = threading.Event()
    eventSaveimage = threading.Event()
    # Detection2 = Detection(devices,root)