python多进程解决多路视频流卡帧、掉帧、延时等问题

多进程解决多路视频流卡帧、掉帧、延时等问题

利用python代码实现不同厂家摄像头同时实时录制并存储

本代码参考: link，文章作者实现读取多个(海康\大华)网络摄像头的视频流 (使用opencv-python)，解决实时读取延迟问题。
本人因项目需求两个摄像机（枪机+球机）同时录制交通监控视频。传统代码可实现单个摄像头实时录制，遇到多视频流会存在通道堵塞，高延时、掉帧、卡帧等诸多问题，在使用link代码时，出现视频流中断的问题。故本文在作者多进程实现多摄像头实时监控的基础上改进视频存放、抓取的代码，实现了多个摄像头实时监测并存储的任务。

本文完整代码如下

import cv2
import time
import multiprocessing as mp


def image_put(q, name, pwd, ip, channel=101):
    cap = cv2.VideoCapture("rtsp://%s:%s@%s//Streaming/Channels/%d" \
                           % (name, pwd, ip, channel))
    fps = cap.get(cv2.CAP_PROP_FPS)
    print('fps: ', fps)
    if cap.isOpened():
        print('HIKVISION1')
    else:
        cap = cv2.VideoCapture("rtsp://%s:%s@%s//Streaming/Channels/%d" \
                               % (name, pwd, ip, channel))
        fps = cap.get(cv2.CAP_PROP_FPS)
        print('fps: ', fps)
        print('HIKVISION2')

    while cap.isOpened():
        # print('cap.read()[0]:', cap.read()[0])
        ret, frame = cap.read()
        # print('ret:', ret)
        frame = cv2.resize(frame, (1920, 1080))
        if not ret:
            cap = cv2.VideoCapture("rtsp://%s:%s@%s//Streaming/Channels/%d" \
                                   % (name, pwd, ip, channel))
            print('HIKVISION2')
            ret, frame = cap.read()
            frame = cv2.resize(frame, (1920,1080))
        q.put(frame)
        # print('q.qsize():', q.qsize())
        q.get() if q.qsize() > 1 else time.sleep(0.01)


def image_get(q, window_name):
    cv2.namedWindow(window_name, cv2.WINDOW_NORMAL)
    fourcc = cv2.VideoWriter_fourcc(*'XVID')
    path = "D:/test/" + window_name + ".avi"
    out = cv2.VideoWriter(path, fourcc, 20.0, (1920, 1080), True)
    while True:
        frame = q.get()
        out.write(frame)
        cv2.imshow(window_name, frame)
        cv2.waitKey(1)


def run_multi_camera():
    # user_name, user_pwd = "admin", "password"
    user_name, user_pwd = "admin", "用户密码"
    camera_ip_l = [
        "192.168.1.67",  # ipv4（改成自己的）
        "192.168.1.64",  # ipv4（改成自己的）
        # 把你的摄像头的地址放到这里，如果是ipv6，那么需要加一个中括号。
    ]

    mp.set_start_method(method='spawn')  # init
    queues = [mp.Queue(maxsize=2) for _ in camera_ip_l]

    processes = []
    for queue, camera_ip in zip(queues, camera_ip_l):
        processes.append(mp.Process(target=image_put, args=(queue, user_name, user_pwd, camera_ip)))
        processes.append(mp.Process(target=image_get, args=(queue, camera_ip)))

    for process in processes:
        process.daemon = True
        process.start()
    for process in processes:
        process.join()


if __name__ == '__main__':
    run_multi_camera()

分模块解读

多进程run_multi_camera()

多进程函数下，将所要使用的设备信息输入对应位置（摄像头可以多个），用户名、密码、设备ip地址自行设置。本人手头两个摄像头，队列参数maxsize=2。
python多进程讲解link

图像放入image_put

不同厂家的摄像头rtsp协议:
video_stream_path = “rtsp://%s:%s@%s/h264/ch%s/main/av_stream” %(user, pwd, ip, channel) # HIKIVISION old version 2015
video_stream_path = “rtsp://%s:%s@%s//Streaming/Channels/%d” % (user, pwd, ip, channel) # HIKIVISION new version 2017
video_stream_path = “rtsp://%s:%s@%s/cam/realmonitor?channel=%d&subtype=0” % (user, pwd, ip, channel) # dahua

本人使用的是海康威视的摄像头

图像提取image_get

设置视频编码格式，设置存储路径（这里我是以摄像头ip地址命名文件）