背景：最近在制作kimera的数据集，尤其是运行semantic模块所需要的bag文件中有很多topic，但是很多不知道topic中装的是什么数据，及其格式，所以我就想着怎么可以将bag中的topic都解析数来，这样就能知道bag中都有啥了．有网友提供了下面的代码，我稍作了些修改，可以在python3下可以顺利运行．如果你的文件夹下有多个bag文件，这个代码可以挨个处理这些bag文件，先为每个bag文件创建一个相应的文件夹，然后把bag文件拷贝进去，然后将bag中的topic都解析为csv格式

 

#!/usr/bin/python
'''
This script saves each topic in a bagfile as a csv.

Accepts a filename as an optional argument. Operates on all bagfiles in current directory if no argument provided

Written by Nick Speal in May 2013 at McGill University's Aerospace Mechatronics Laboratory
www.speal.ca

Supervised by Professor Inna Sharf, Professor Meyer Nahon

'''

import rosbag, sys, csv
import time
import string
import os #for file management make directory
import shutil #for file management, copy file

#verify correct input arguments: 1 or 2
if (len(sys.argv) > 2):
	print("invalid number of arguments:   " + str(len(sys.argv)))
	print("should be 2: 'bag2csv.py' and 'bagName'")
	print("or just 1  : 'bag2csv.py'")
	sys.exit(1)
elif (len(sys.argv) == 2):
	listOfBagFiles = [sys.argv[1]]
	numberOfFiles = "1"
	print("reading only 1 bagfile: " + str(listOfBagFiles[0]))
elif (len(sys.argv) == 1):
	listOfBagFiles = [f for f in os.listdir(".") if f[-4:] == ".bag"]	#get list of only bag files in current dir.
	numberOfFiles = str(len(listOfBagFiles))
	print("reading all " + numberOfFiles + " bagfiles in current directory: \n")
	for f in listOfBagFiles:
		print(f)
	print("\n press ctrl+c in the next 10 seconds to cancel \n")
	time.sleep(10)
else:
	print("bad argument(s): " + str(sys.argv))	#shouldnt really come up
	sys.exit(1)

count = 0
for bagFile in listOfBagFiles:
	count += 1
	print("reading file " + str(count) + " of  " + numberOfFiles + ": " + bagFile)
	#access bag
	bag = rosbag.Bag(bagFile)
	bagContents = bag.read_messages()
	bagName = bag.filename#/home/yunlei/COOL/kalibr-cde/test/2020-06-13-11-57-29.bag

	#create a new directory string.rstrip(bagName, ".bag")
	folder = bagName.split(".bag")[0]
	try:	#else already exists
		os.makedirs(folder)
	except:
		pass
	shutil.copyfile(bagName, folder + '/' + bagName.split('/')[-1])


	#get list of topics from the bag
	listOfTopics = []
	for topic, msg, t in bagContents:
		if topic not in listOfTopics:
			listOfTopics.append(topic)


	for topicName in listOfTopics:
		#Create a new CSV file for each topic folder + '/' + string.replace(topicName, '/', '_slash_') + '.csv'
		filename = folder + '/' + topicName.split('/')[-1] + '.csv'
		with open(filename, 'w+') as csvfile:
			filewriter = csv.writer(csvfile, delimiter = ',')
			firstIteration = True	#allows header row
			for subtopic, msg, t in bag.read_messages(topicName):	# for each instant in time that has data for topicName
				#parse data from this instant, which is of the form of multiple lines of "Name: value\n"
				#	- put it in the form of a list of 2-element lists
				msgString = str(msg)
				msgList = msgString.split('\n')
				instantaneousListOfData = []
				for nameValuePair in msgList:
					splitPair = nameValuePair.split(':')
					for i in range(len(splitPair)):	#should be 0 to 1
						splitPair[i] = splitPair[i].strip()
					instantaneousListOfData.append(splitPair)
				#write the first row from the first element of each pair
				if firstIteration:	# header
					headers = ["rosbagTimestamp"]	#first column header
					for pair in instantaneousListOfData:
						headers.append(pair[0])
					filewriter.writerow(headers)
					firstIteration = False
				# write the value from each pair to the file
				values = [str(t)]	#first column will have rosbag timestamp
				for pair in instantaneousListOfData:
					if len(pair) > 1:
						values.append(pair[1])
				filewriter.writerow(values)
	bag.close()
print("Done reading all " + numberOfFiles + " bag files.")

 

rostopic pub -1 /turtle1/cmd_vel geometry_msgs/Twist -- '[2,0,0]' '[0,0,1.8]'

 

 

https://www.cnblogs.com/BlueMountain-HaggenDazs/p/6277100.html

 

map与costmap都是以nav_msgs::OccupancyGrid类型发布其topic。其中整张地图的障碍物信息存放在data数据成员中，data是一个int8类型的vector向量，即一维数组。假设一张pgm的map地图，宽：width，高：height，单位为像素，分辨路为resolution，左下角像素点在世界坐标系下的位置为：（ox,oy）,单位米，那么世界坐标系下一点（x,y）单位米，假设其在地图中，那么该点对应的data中的索引index为：

index = (x-ox)/resolution+(y-oy)/resolution*width

（注：index为int8类型，所以要将结果进行下类型转换），

那么该点在地图中的信息即为data[index]

即：data是按照那张地图图片的自底向上，自左至右逐个像素点存储的

map中data存储的格式如下：

0:空白区域

100:障碍物

-1:未知

1-99:根据像素的灰度值转换后的值，代表有障碍物的概率（不确定，猜测是这个含义，实际用的时候把这个值都当作无障碍物）

costmap中data存储格式如下：

0:空白区域

100:障碍物层

99:膨胀层

1-98:距离障碍物层的远近获得的代价值

-1:未知区域

node发布的topic格式   /+名称（可以是二级名称的形式）
发布的topic放在master中，所有的node都可以订阅，但是名称要一样
当一个node订阅一个topic，但是这个topic的node还没有启动，这时rosnode info 这个node，就可以看到他所订阅的这个topic还是未定义类型。