【python】使用ffmpeg+opencv读取摄像头并推流到rtmp服务器
1、环境
python3
OSX 12.5
vscode
2、安装ffmpeg
brew install ffmpeg
3、安装cv2
pip install opencv-python # or pip --default-timeout=100 install opencv-python -i https://pypi.douban.com/simple
4、脚本
import cv2 # subprocess 模块允许我们启动一个新进程,并连接到它们的输入/输出/错误管道,从而获取返回值。 import subprocess # 视频读取对象 cap = cv2.VideoCapture(0) # 推流地址 rtmp = "rtmp://192.168.10.225:1935/stream/example"# 推流的服务器地址 # 设置推流的参数 command = ['ffmpeg', '-y', '-f', 'rawvideo', '-vcodec', 'rawvideo', '-pix_fmt', 'bgr24', '-s', '1280*720', # 根据输入视频尺寸填写 '-r', '25', '-i', '-', '-c:v', 'h264', '-pix_fmt', 'yuv420p', '-preset', 'ultrafast', '-f', 'flv', rtmp] # 创建、管理子进程 pipe = subprocess.Popen(command, stdin=subprocess.PIPE) size = (int(cap.get(cv2.CAP_PROP_FRAME_WIDTH)), int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))) # 循环读取 while cap.isOpened(): # 读取一帧 ret, frame = cap.read() if frame is None: print('read frame err!') continue # 显示一帧 # fps = int(cap.get(cv2.CAP_PROP_FPS)) cv2.imshow("frame", frame) # 按键退出 if cv2.waitKey(1) & 0xFF == ord('q'): break # 读取尺寸、推流 img = cv2.resize(frame, size) pipe.stdin.write(img.tobytes()) # 关闭窗口 cv2.destroyAllWindows() # 停止读取 cap.release()
备注:分辨率要根据本机摄像头支持的分比率设置,不然可能显示不出图像
参考链接:
http://events.jianshu.io/p/4d267a9267a5
https://blog.csdn.net/qq_41876456/article/details/119480972
https://blog.csdn.net/m0_51573312/article/details/124580266