使用ros订阅图像话题并保存成图像

下载代码后，修改:

conda环境
视频保存路径
图像话题

使用：运行后，在终端输入start并回车开始录制，输入stop并回车结束录制。

#!/home/cvpr/anaconda3/envs/ros_py37/bin/python3.7
# -*- coding: utf-8 -*-"""
订阅图像并保存为视频
"""import rospy
import cv2
from sensor_msgs.msg import Image
from cv_bridge import CvBridgeclass VideoRecorder:def __init__(self, path):self.bridge = CvBridge()self.image_sub = rospy.Subscriber('/rgb/image_raw', Image, self.image_callback)self.frames = []self.recording = Falseself.path = pathdef image_callback(self, msg):try:cv_image = self.bridge.imgmsg_to_cv2(msg, "bgr8")if self.recording:self.frames.append(cv_image)except Exception as e:print(e)def start_recording(self):self.frames = []self.recording = Truedef stop_recording(self):self.recording = Falseif self.frames:self.save_video()def save_video(self):height = self.frames[0].shape[0]weight = self.frames[0].shape[1]fps = 30# fourcc = cv.VideoWriter_fourcc('M', 'J', 'P', 'G') 用于avi格式的生成fourcc = cv2.VideoWriter_fourcc(*'mp4v')  # 用于mp4格式的生成videowriter = cv2.VideoWriter(self.path, fourcc, fps, (weight, height))  # 创建一个写入视频对象for img in self.frames:videowriter.write(img)videowriter.release()print('save to', self.path)if __name__ == '__main__':rospy.init_node('video_recorder_node', anonymous=True)path = 'new.mp4'recorder = VideoRecorder(path)try:while not rospy.is_shutdown():cmd = input("Enter 'start' to begin recording or 'stop' to stop recording: ")if cmd == 'start':recorder.start_recording()elif cmd == 'stop':recorder.stop_recording()except rospy.ROSInterruptException:pass