利用ROS工具從bag包中提取圖片和.csv文件

1.查看bag文件中topic、內容：

rosbag info xx.bag

2.安裝圖片處理工具

sudo apt-get install ffmpeg

3. 新建存儲圖片的文件夾cam0，執行如下命令：

rosrun image_view extract_images _sec_per_frame:=0.01 image:=/mynteye/left/image_mono

這種方式提取的圖片的名稱是從0依次排序的，若想提取為圖片名是時間戳的需要手寫一個.py文件，利用python回放。下面是我的.py文件。

# coding:utf-8
#!/usr/bin/python
 
# Extract images from a bag file.
 
#PKG = 'beginner_tutorials'
import roslib   #roslib.load_manifest(PKG)
import rosbag
import rospy
import decimal
import cv2
from sensor_msgs.msg import Image
from cv_bridge import CvBridge
from cv_bridge import CvBridgeError
 
# Reading bag filename from command line or roslaunch parameter.
#import os
#import sys
 
camera0_path = '/home/ubuntu/code/dataset/loop/cam0/data/'
camera1_path = '/home/ubuntu/code/dataset/loop/cam1/data/'
 
class ImageCreator():
 
 
    def __init__(self):
        self.bridge = CvBridge()
        with rosbag.Bag('/home/ubuntu/code/dataset/looppp/loop.bag', 'r') as bag:  #要讀取的bag文件；
            for topic,msg,t in bag.read_messages():
                if topic == "/mynteye/left/image_mono": #圖像的topic；
                        try:
                            cv_image = self.bridge.imgmsg_to_cv2(msg,"bgr8")
                        except CvBridgeError as e:
                            print e
                        timestr = msg.header.stamp.to_sec()
                        #%.6f表示小數點后帶有6位，可根據精確度需要修改；
                        timer = 100000000 * timestr
                        image_name = "%d" % timer + ".png" #圖像命名：時間戳.png
                        cv2.imwrite(camera0_path + image_name, cv_image)  #保存；
                elif topic == "/mynteye/right/image_mono": #圖像的topic；
                        try:
                            cv_image = self.bridge.imgmsg_to_cv2(msg,"bgr8")
                        except CvBridgeError as e:
                            print e
                        timestr = msg.header.stamp.to_sec()
                        #%.6f表示小數點后帶有6位，可根據精確度需要修改；
                        timer = 100000000 * timestr
                        image_name = "%d" % timer + ".png" #圖像命名：時間戳.png
                        cv2.imwrite(camera1_path + image_name, cv_image)  #保存；
 
if __name__ == '__main__':
 
    #rospy.init_node(PKG)
 
    try:
        image_creator = ImageCreator()
    except rospy.ROSInterruptException:
        pass

4. 打開.bag文件所在位置，執行如下命令：

rosbag play xx.bag

5.將非圖片數據提取為.csv格式存儲：

rostopic echo -b xx.bag -p /mynteye/left/image_mono > cam0data.csv
rostopic echo -b xx.bag -p /mynteye/right/image_mono > cam1data.csv
rostopic echo -b xx.bag -p /mynteye/imu/imu_raw > imudata.csv