openCV 多线程调用多个海康威视摄像头（避免画面不同步）,进行人脸识别

1、创建多个线程来拉取不同的视频流，不使用多线程的话，所有的视频流就在排一个队，会出现画面不同步的情况

代码：

from threading import Thread
import cv2


#视频流拉取类
class vStream:
    def __init__(self,src):
        self.capture = cv2.VideoCapture(src)
        self.thread = Thread(target=(self.update),args=())
        self.thread.daemon = True
        self.thread.start()
    def update(self):
        while True:
            _,self.frame = self.capture.read()
    def getFrame(self):
        return self.frame


#我的网络摄像头
cam1 = vStream('rtsp://admin:qwertyuiop123@192.168.31.103/Streaming/Channels/2')
cam2 = vStream('rtsp://admin:JGSYS123@192.168.31.123/Streaming/Channels/2')

while True:
    try:
        myframe1 = cam1.getFrame()
        myframe2 = cam2.getFrame()
        cv2.imshow('webCam1',myframe1)
        cv2.imshow('webCam2',myframe2)
    except:
        print('frame not available')

    if cv2.waitKey(1) ==ord('q'):
        cam1.capture.release()
        cam2.capture.release()
        cv2.destroyAllWindows()
        exit(1)
        break

2、使用numpy将多个画面放在同一个frame中，图像就是矩阵，numpy是用来处理矩阵的类

代码：

import time
from threading import Thread
import cv2
import numpy as np


#视频流拉取类
class vStream:
    def __init__(self,src,width,height):
        #frame 初始尺寸
        self.width = width
        self.height = height
        self.capture = cv2.VideoCapture(src)
        self.thread = Thread(target=(self.update),args=())
        self.thread.daemon = True
        self.thread.start()
    def update(self):
        while True:
            _,self.frame = self.capture.read()
            self.frame2 = cv2.resize(self.frame,(self.width,self.height))

    def getFrame(self):
        return self.frame2

ip1 = 'rtsp://admin:qwertyuiop123@192.168.31.103/Streaming/Channels/2'
ip2 = 'rtsp://admin:JGSYS123@192.168.31.123/Streaming/Channels/2'

disW = 640
disH = 360
cam1 = vStream(ip1,disW,disH)
cam2 = vStream(ip2,disW,disH)
#显示帧率
font = cv2.FONT_HERSHEY_SIMPLEX
startTime = time.time()
dtav = 1#没啥意义，只是为了第一次进入循环时没有dtav避免错误

while True:
    try:
        myframe1 = cam1.getFrame()
        myframe2 = cam2.getFrame()
        #将myframe1，myframe2 放在同一个画面
        frameCombine = np.hstack((myframe1,myframe2))

        #显示帧率
        dt = time.time() - startTime
        startTime = time.time()
        # 低通滤波，避免fps误差太大
        dtav = .90 * dtav + .1 * dt
        fps = 1 / dtav
        fps = round(fps, 2)
        # print('fps is:',round(dtav,1))
        cv2.rectangle(frameCombine, (0, 0), (100, 40), (0, 0, 255), -1)
        cv2.putText(frameCombine, str(round(fps, 1)) + 'fps', (0, 25), font, .75, (0, 255, 255, 2))
        cv2.imshow('combine frame', frameCombine)

    except:
        print('frame not available')

    keyCode = cv2.waitKey(1) & 0xFF
    if keyCode == 27:  # ESC键退出
        cam1.capture.release()
        cam2.capture.release()
        cv2.destroyAllWindows()
        exit(1)
        break

 进行人脸识别，网不好帧率太低

import face_recognition
import pickle
import time
from threading import Thread
import cv2
import numpy as np

with open('train.pkl','rb')as f:
    Names = pickle.load(f)
    Encodings = pickle.load(f)



#视频流拉取类
class vStream:
    def __init__(self,src,width,height):
        #frame 初始尺寸
        self.width = width
        self.height = height
        self.capture = cv2.VideoCapture(src)
        self.thread = Thread(target=(self.update),args=())
        self.thread.daemon = True
        self.thread.start()
    def update(self):
        while True:
            _,self.frame = self.capture.read()
            self.frame2 = cv2.resize(self.frame,(self.width,self.height))

    def getFrame(self):
        return self.frame2

ip1 = 'rtsp://admin:qwertyuiop123@192.168.31.103/Streaming/Channels/2'
ip2 = 'rtsp://admin:JGSYS123@192.168.31.123/Streaming/Channels/2'

disW = 640
disH = 360
cam1 = vStream(ip1,disW,disH)
cam2 = vStream(ip2,disW,disH)
#显示帧率
font = cv2.FONT_HERSHEY_SIMPLEX
startTime = time.time()
dtav = 1
scaleFactor=.25

while True:
    try:
        myframe1 = cam1.getFrame()
        myframe2 = cam2.getFrame()
        #将myframe1，myframe2 放在同一个画面
        frameCombine = np.hstack((myframe1,myframe2))
        #将图像转换缩小，加载模型识别
        frameRGB = cv2.cvtColor(frameCombine, cv2.COLOR_BGR2RGB)
        frameRGBSmall = cv2.resize(frameRGB, (0, 0), fx=scaleFactor, fy=scaleFactor)
        facePositions = face_recognition.face_locations(frameRGBSmall, model='cnn')
        allEncodings = face_recognition.face_encodings(frameRGBSmall, facePositions)

        for (top, right, bottom, left), face_encoding in zip(facePositions, allEncodings):
            name = 'Unkown Person'
            matches = face_recognition.compare_faces(Encodings, face_encoding)
            if True in matches:
                first_match_index = matches.index(True)
                name = Names[first_match_index]
                # 将检测框放大scalefactor倍，上面图片是缩小三倍检测的结果，检测框也是缩小三倍的
            top = int(top / scaleFactor)
            right = int(right / scaleFactor)
            bottom = int(bottom / scaleFactor)
            left = int(left / scaleFactor)
            cv2.rectangle(frameCombine, (left, top), (right, bottom), (0, 0, 255), 2)
            cv2.putText(frameCombine, name, (left, top - 6), font, .75, (0, 0, 255), 2)

        #显示帧率
        dt = time.time() - startTime
        startTime = time.time()
        # 低通滤波，避免fps误差太大
        dtav = .90 * dtav + .1 * dt
        fps = 1 / dtav
        fps = round(fps, 2)
        # print('fps is:',round(dtav,1))
        cv2.rectangle(frameCombine, (0, 0), (100, 40), (0, 0, 255), -1)
        cv2.putText(frameCombine, str(round(fps, 1)) + 'fps', (0, 25), font, .75, (0, 255, 255, 2))
        cv2.imshow('combine frame', frameCombine)

    except:
        print('frame not available')

    keyCode = cv2.waitKey(1) & 0xFF
    if keyCode == 27:  # ESC键退出
        cam1.capture.release()
        cam2.capture.release()
        cv2.destroyAllWindows()
        exit(1)
        break