Table of contents
1. Background
It is necessary to use pyqt5 to make a simple player, which can connect to the unit camera, use rtsp to pull the stream and display it in real time. Many blogging methods have been tried, or they are stuck, or the delay is too large, which cannot meet the needs.
2. Realize
Mainly refer to the blog "PyQt multi-threaded video playback (QThread, QRunnable, moveToThread three thread playback methods)" , this blog mainly uses two threads, thread 1 is used to pull streams and decode, thread 2 is used to process pictures, process Then it is displayed on the qlabel() of the main thread. A global queue is used between thread 1 and thread 2 to exchange picture frames.
Different from the above blog, using the global queue will cause serious delay, so it is replaced by the global stack. At the same time, after thread 2 obtains the latest video frame from the global stack, the global stack needs to be cleared to ensure that thread 2 gets the latest video frame every time. All are the latest picture frames.
code show as below:
# coding=gbk //放到文件首行
import sys,math
from PyQt5.QtGui import *
from PyQt5.QtWidgets import *
from PyQt5.QtCore import *
import numpy as np
import cv2
import os
import time
import base64
import copy
from PIL import Image, ImageDraw, ImageFont
from queue import LifoQueue
import threading
Decode2Play = LifoQueue()
#lock = threading.Lock()
class cvDecode(QThread):
def __init__(self):
super(cvDecode, self).__init__()
self.threadFlag = 0 # 控制线程退出
self.rtsp = "rtsp://用户名:密码@摄像头ip"
self.cap = cv2.VideoCapture(self.rtsp)
def run(self):
#print("当前线程cvDecode: self.threadFlag:{}".format(self.threadFlag))
while self.threadFlag:
if self.cap.isOpened():
ret, frame = self.cap.read()
time.sleep(0.001) # 控制读取录像的时间，连实时视频的时候改成time.sleep(0.001)，多线程的情况下最好加上，否则不同线程间容易抢占资源
if ret:
#lock.acquire()
Decode2Play.put(frame) # 解码后的数据放到队列中
#lock.release()
del frame # 释放资源
class play_Work(QThread):
def __init__(self):
super(play_Work, self).__init__()
self.threadFlag = 0 # 控制线程退出
self.playLabel = QLabel() # 初始化QLabel对象
#cv2.namedWindow("test")
#cv2.resizeWindow("test", 640, 480)
def run(self):
while self.threadFlag:
if not Decode2Play.empty():
frame = Decode2Play.get()
while not Decode2Play.empty():
Decode2Play.get()
frame = cv2.resize(frame, (800, 600), cv2.INTER_LINEAR)
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
qimg = QImage(frame.data, frame.shape[1], frame.shape[0], QImage.Format_RGB888) # 在这里可以对每帧图像进行处理，
self.playLabel.setPixmap(QPixmap.fromImage(qimg)) # 图像在QLabel上展示
#cv2.imshow('test', frame)
time.sleep(0.001)
class MyWindow(QWidget):
#opencam_complete_signal = pyqtSignal(str) # 子线程返回摄像头图像到主线程的信号
def __init__(self):
super().__init__()
self.setGeometry (0, 0, 840, 800)
#self.setFixedSize(1600, 960)
self.init_ui()
def init_ui(self):
self.cam_view = QLabel(self)
self.cam_view.setStyleSheet("border:2px solid black")
self.cam_view.setGeometry(20, 20, 800, 600)
self.open_cam_btn = QPushButton(self)
#self.pushButton.setStyleSheet("border:2px solid black")
self.open_cam_btn.setText('播放')
self.open_cam_btn.setGeometry(380, 700, 100, 30)
self.open_cam_btn.clicked.connect(self.init_work)
def init_work(self):
self.decodework = cvDecode()
self.decodework.threadFlag = 1
self.decodework.start()
self.playwork = play_Work()
self.playwork.threadFlag = 1
self.playwork.playLabel = self.cam_view
self.playwork.start()
def closeEvent(self, event):
print("关闭线程")
# Qt需要先退出循环才能关闭线程
if self.decodework.isRunning():
self.decodework.threadFlag = 0
self.decodework.quit()
if self.playwork.isRunning():
self.playwork.threadFlag = 0
self.playwork.quit()
if __name__ == '__main__':
app = QApplication(sys.argv) # 创建对象
w = MyWindow()
# 展示窗口
w.show()
# 程序进行循环等待状态
app.exec_()
The effect is as follows:
