0. 前言

我们都知道ROS1与ROS2的通信是不一样的，而ROS2也因为DDS的通讯带来了一些问题，其中最严重的就属于在存储大数据包的时候会出现的丢帧问题。而如何解决这样的问题目前官方没有给出非常好的解决方案，所以我们只有曲线救国，通过保存图片的方式来实现bag包的完整录制。
操作系统:
Ubuntu 20.04
版本：foxy
DDS实现：Fast-RTPS
客户端库(如适用)：rclcpp/rclpy

1. 问题说明

我们在使用ROS1与ROS2的bag互转时发现会存在丢帧的情况，同时我们直接在ROS2中录制bag包也会发现录制的频率是存在问题的

ROS2

Files:             rosbag2_2021_05_19-11_28_34/rosbag2_2021_05_19-11_28_34_0.db3
Bag size:          1.1 GiB
Storage id:        sqlite3
Duration:          42.616s
Start:             May 19 2021 11:28:34.289 (1621394914.289)
End                May 19 2021 11:29:16.906 (1621394956.906)
Messages:          857
Topic information: Topic: /image | Type: sensor_msgs/msg/Image | Count: 427 | Serialization Format: cdr
                   Topic: /scan | Type: sensor_msgs/msg/LaserScan | Count: 430 | Serialization Format: cdr

ROS1

path:        2021-05-19-19-59-37.bag
version:     2.0
duration:    42.5s
start:       May 19 2021 19:59:41.91 (1621425581.91)
end:         May 19 2021 20:00:24.38 (1621425624.38)
size:        1.1 GB
messages:    847
compression: none [419/419 chunks]
types:       sensor_msgs/Image     [060021388200f6f0f447d0fcd9c64743]
             sensor_msgs/LaserScan [90c7ef2dc6895d81024acba2ac42f369]
topics:      /image   418 msgs    : sensor_msgs/Image    
             /scan    429 msgs    : sensor_msgs/LaserScan

这里经过测试得到了下面的数据

在PC端：
在这里插入图片描述
在Xavier端
在这里插入图片描述
我们可以发现相较于ROS1稳定而言，ROS2的频率下降很严重，并不能拉满宽带。原本以为是CPU问题，但是经过测试发现CPU变化不大，个人怀疑是DDS通讯在ROS2中本身造成的问题。在ROS官网中也发现了该问题的阐述，原因是因为FastRTPS这类DDS并不能传输较大的带宽。我们可以更换DDS来缓解这样的问题，但是仍然存在丢帧的问题。为此本文提供了一个不走ros_bridge1方式的解决方案
在这里插入图片描述

2.无丢帧转bag方式

首先从ROS1中订阅bag包或者摄像头的数据，并将对应的图片以时间戳的形式保存。

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import rospy
from geometry_msgs.msg import Twist

import  select, termios, tty
from sensor_msgs.msg import Image
from cv_bridge import CvBridge, CvBridgeError

import cv2
import numpy as np
bridge = CvBridge()
img=[]
def vels(speed,turn):
    return "currently:\tspeed %s\tturn %s " % (speed,turn)

def image_callback(msg):
    print("Received an image!")
        # Convert your ROS Image message to OpenCV
     try:
         cv_image = self.bridge.imgmsg_to_cv2(msg,"bgr8")
     except CvBridgeError as e:
         print e
     timestr = "%.6f" %  msg.header.stamp.to_sec()
     #%.6f表示小数点后带有6位，可根据精确度需要修改；
     image_name = timestr+ ".jpg" #图像命名：时间戳.jpg
     cv2.imwrite(image_name, cv_image)  #保存；

if __name__=="__main__":

    rospy.init_node('mbot_teleop')
    pub = rospy.Publisher('/cmd_vel', Twist, queue_size=5)
    rospy.Subscriber('/camera/image_raw',Image,image_callback)
    rospy.spin()

ROS2部分读取对应的文件夹，并将对应的文件publish出来，并通过ros2 bag record的方式录制

import cv2
import rclpy
from rclpy.node import Node
from sensor_msgs.msg import Image
from cv_bridge import CvBridge
from cv_bridge import CvBridge, CvBridgeError

class Image_Publisher(Node):
    def __init__(self):
        super().__init__("image_tutorial")
        self.add_msg_to_info_logger("initializing node")
        self.image_publisher_ = self.create_publisher(Image, 'image', 10)
        self.bridge = CvBridge()


    def publish(self, cv2_image):
        try:
             msg = self.bridge.cv2_to_imgmsg(cv2_image, "bgr8")
             self.image_publisher_.publish(msg)
        except CvBridgeError as e:
             self.add_msg_to_info_logger(e)
        self.add_msg_to_info_logger("sending_image")


    def add_msg_to_info_logger(self, msg):
        self.get_logger().info(msg)

rclpy.init()
image_publisher = Image_Publisher()

image_publisher.add_msg_to_info_logger("initializing camera")
path="xxxx/xxxxx/"
for file in os.listdir(path):
    file_path = os.path.join(path, file)
    if os.path.isdir(file_path):
        listdir(file_path, list_name)
    else:
        list_name.append(file_path)
len_list = len(list_name)
cv2.namedWindow("image_tutorial")

for i in len_list:
    frame = cv2.imread(i)//读取图片
    cv2.imshow("image_tutorial", frame)
    pressed = cv2.waitKey(1)
    if pressed == 27 or pressed == ord('q'):
    # closing application if esc or q pressed
        image_publisher.add_msg_to_info_logger("closing button pressed, closing")
        break

    elif pressed == 32:
    # publishing image if space pressed
        image_publisher.publish(frame)

# closing  
image_publisher.destroy_node()
rclpy.shutdown()   
cam.release()
cv2.destroyAllWindows()