Cartographer使用３D激光雷达建立２D导航图（概率栅格地图）_cartographer 2d定位可以导入3d模式下建好的地图吗

网上学习资料一大堆，但如果学到的知识不成体系，遇到问题时只是浅尝辄止，不再深入研究，那么很难做到真正的技术提升。

需要这份系统化的资料的朋友，可以添加戳这里获取

一个人可以走的很快，但一群人才能走的更远！不论你是正从事IT行业的老鸟或是对IT行业感兴趣的新人，都欢迎加入我们的的圈子（技术交流、学习资源、职场吐槽、大厂内推、面试辅导），让我们一起学习成长！

<launch>
  <param name="robot\_description"
    textfile="$(find cartographer\_ros)/urdf/my\_backpack\_3d.urdf" />//换成自己的机器人配置文件

  <node name="robot\_state\_publisher" pkg="robot\_state\_publisher"
    type="robot\_state\_publisher" />

  <node name="cartographer\_node" pkg="cartographer\_ros"
      type="cartographer\_node" args="
          -configuration_directory $(find cartographer_ros)/configuration_files
          -configuration_basename my_robot_3d.lua"	//换成自己配置文件
      output="screen">
	<remap from="points2" to="/rfans\_driver/rfans\_points" />//换成自己的bag包数据中的点云话题
	
  </node>

  <node name="cartographer\_occupancy\_grid\_node" pkg="cartographer\_ros"
      type="cartographer\_occupancy\_grid\_node" args="-resolution 0.05" />  
</launch>

会看点云话题吗？用这个命令查看

# rosbag info \*\*\*\*\*.bag

看到这个topics中sensor_msgs/PointCloud2的这个话题类型了吗？前面的就是你自己数据的点云话题，我的点云话题是/rfans_driver/rfans_points,所以在launch文件中，把点云话题改成自己的

3. 配置lua文件
首先看一下使用官方backpack_3d.lua文件生成图的效果

include "map\_builder.lua"
include "trajectory\_builder.lua"

options = {
  map_builder = MAP\_BUILDER,
  trajectory_builder = TRAJECTORY\_BUILDER,
  map_frame = "map",
  tracking_frame = "base\_link",
  published_frame = "base\_link",
  odom_frame = "odom",
  provide_odom_frame = true,
  publish_frame_projected_to_2d = false,
  use_odometry = false,
  use_nav_sat = false,
  use_landmarks = false,
  num_laser_scans = 0,
  num_multi_echo_laser_scans = 0,
  num_subdivisions_per_laser_scan = 1,
  num_point_clouds = 2,
  lookup_transform_timeout_sec = 0.2,
  submap_publish_period_sec = 0.3,
  pose_publish_period_sec = 5e-3,
  trajectory_publish_period_sec = 30e-3,
  rangefinder_sampling_ratio = 1.,
  odometry_sampling_ratio = 1.,
  fixed_frame_pose_sampling_ratio = 1.,
  imu_sampling_ratio = 1.,
  landmarks_sampling_ratio = 1.,
}

TRAJECTORY\_BUILDER\_3D.num_accumulated_range_data = 160

MAP\_BUILDER.use_trajectory_builder_3d = true
MAP\_BUILDER.num_background_threads = 7
POSE\_GRAPH.optimization_problem.huber_scale = 5e2
POSE\_GRAPH.optimize_every_n_nodes = 320
POSE\_GRAPH.constraint_builder.sampling_ratio = 0.03
POSE\_GRAPH.optimization_problem.ceres_solver_options.max_num_iterations = 10
POSE\_GRAPH.constraint_builder.min_score = 0.62
POSE\_GRAPH.constraint_builder.global_localization_min_score = 0.66

return options

看一下官方3d bag包跑出来的效果
看起来效果还可以吧
再看一下我建的图
我先说一下这张图的问题：１．为什么没有出现白色区域（已知区域）２．为什么有这么多的噪点
后来重新调整了lua文件
my_robot_3d.lua

include "map\_builder.lua"
include "trajectory\_builder.lua"


options = {
  map_builder = MAP\_BUILDER,
  trajectory_builder = TRAJECTORY\_BUILDER,
  map_frame = "map",
  tracking_frame = "base\_link",	--imu
  published_frame = "base\_link",   --laser
  odom_frame = "odom",
  provide_odom_frame = false,
  publish_frame_projected_to_2d = false,
  use_odometry = false,
  use_nav_sat = false,
  use_landmarks = false,
  num_laser_scans =0,
  num_multi_echo_laser_scans = 0,
  num_subdivisions_per_laser_scan = 1,
  num_point_clouds = 1,
  lookup_transform_timeout_sec = 0.2,
  submap_publish_period_sec = 0.3,
  pose_publish_period_sec = 5e-3,
  trajectory_publish_period_sec = 30e-3,
  rangefinder_sampling_ratio = 1.,
  odometry_sampling_ratio = 1,    --1.
  fixed_frame_pose_sampling_ratio = 1.,
  imu_sampling_ratio = 1.,
  landmarks_sampling_ratio = 1.,
}

TRAJECTORY\_BUILDER\_2D.min_z = 0.1
TRAJECTORY\_BUILDER\_2D.max_z = 1.5
TRAJECTORY\_BUILDER\_3D.num_accumulated_range_data = 1   	--1
TRAJECTORY\_BUILDER\_3D.min_range = 1.3			--houjia
TRAJECTORY\_BUILDER\_3D.max_range = 25.			--houjia
TRAJECTORY\_BUILDER\_3D.use_online_correlative_scan_matching = true		--houjia
--TRAJECTORY\_BUILDER\_3D.submaps.num_range_data=90.

MAP\_BUILDER.use_trajectory_builder_3d = true
MAP\_BUILDER.num_background_threads = 7
POSE\_GRAPH.optimization_problem.huber_scale = 5e2
POSE\_GRAPH.optimize_every_n_nodes = 60
POSE\_GRAPH.constraint_builder.sampling_ratio = 0.03
POSE\_GRAPH.optimization_problem.ceres_solver_options.max_num_iterations = 10
POSE\_GRAPH.constraint_builder.min_score = 0.62
POSE\_GRAPH.constraint_builder.global_localization_min_score = 0.66

return options

出现了如下效果图，噪点没有了，至关重要的一个参数是TRAJECTORY_BUILDER_3D.use_online_correlative_scan_matching = true
这个参数的作用就是开启实时的闭环检测来进行前段的扫描匹配，所以生成的图效果非常好，就是计算量会增大，消耗更多CPU。

但是这个图还是不能用来导航啊，只有灰色和黑色区域（也就是障碍物和未知区域）。

现在剩下的问题就是要解决白色区域了

1.2在线或离线生成pbstream文件，再转成pgm，yaml文件

在线生成pbstream文件，还是启动上个方法的launch文件

# roslaunch cartographer_ros my_demo_backpack_3d.launch bag_filename:=${HOME}/Downloads/\*\*\*.bag

在rviz生成图了之后，启动一个新终端使用如下命令：

＃ rosservice call /finish_trajectory 0　	//停止接受传感器数据
＃ rosservice call /write_state  /home/路径/\*\*\*.pbstream //生成pbstream文件

接下来就是生成pgm文件

＃ roslaunch cartographer_ros assets_writer_ros_map.launch bag_filenames:=${HOME}/Downloads/2020-04-29-11-25-22.bag pose_graph_filename:=${HOME}/Downloads/2020-04-29-11-25-22.bag.pbstream

把bag文件路径和pbstream文件路径改成你自己的，生成图如下：

但是在路线闭环的情况下，就会出现激光雷达的波纹信息，如下图

虽然这种方法有白色可通行区域，但这个雷达波纹和行车轨迹看起来很恶心啊。
离线的方法我就不说了，效果是一样的。离线生成启动的launch文件是

roslaunch cartographer_ros offline_backpack_3d.launch bag_filenames:=${HOME}/Downloads/2020-05-13-16-37-40.bag

一样把launch文件换成自己的点云话题！运行完之后会直接生成pbstream文件

２.最终方案！使用３d点云数据直接在rviz上生成２d图（导航图，栅格图）

这种方法是需要里程计/odom信息，不用imu信息
先贴自己配置图
demo_zch_2d.launch

<launch>
  <param name="/use\_sim\_time" value="true" />
  <!-- <node pkg="tf" type="static\_transform\_publisher" name="base\_laser\_broadcaster" args="0 0 0 0 0 0 laserbase\_footprint /base\_footprint 20" /> -->
  <include file="$(find cartographer\_ros)/launch/zch\_rslidar\_2d.launch" />
  <node name="cartographer\_node" pkg="cartographer\_ros"
      type="cartographer\_node" args="
          -configuration_directory
              $(find cartographer_ros)/configuration_files
          -configuration_basename zch_rslidar_2d.lua"//你自己的lua文件
      output="screen">
      <remap from="/odom" to="/odom" />		//你自己的里程计话题
      <remap from="/points2" to="/rfans\_driver/rfans\_points" />　　//你自己的点云话题
  </node>
  <node name="rviz" pkg="rviz" type="rviz" required="true"
      args="-d $(find cartographer\_ros)/configuration\_files/rslidar\_2d.rviz" />
  <node name="cartographer\_occupancy\_grid\_node" pkg="cartographer\_ros"
      type="cartographer\_occupancy\_grid\_node" args="-resolution 0.05" />
</launch>

zch_rslidar_2d.launch

<launch>
  <param name="robot\_description"
    textfile="$(find cartographer\_ros)/urdf/my\_backpack\_3d.urdf" />

  <node name="robot\_state\_publisher" pkg="robot\_state\_publisher"
    type="robot\_state\_publisher" />
</launch>

zch_rslidar_2d.lua

include "map\_builder.lua"
include "trajectory\_builder.lua"

options = {
  map_builder = MAP\_BUILDER,
  trajectory_builder = TRAJECTORY\_BUILDER,
  map_frame = "map",
  tracking_frame = "base\_link",
  published_frame = "base\_link",
  odom_frame = "odom",
  provide_odom_frame = true,
  publish_frame_projected_to_2d = false,
  use_odometry = true,
  use_nav_sat = false,
  use_landmarks = false,
  num_laser_scans = 0,
  num_multi_echo_laser_scans = 0,
  num_subdivisions_per_laser_scan = 1,


![img](https://img-blog.csdnimg.cn/img_convert/ad8a79cf30dc0ac7216b8ea9d7fcd888.png)
![img](https://img-blog.csdnimg.cn/img_convert/c2ece18a977883f132dc53ad92b72d4a.png)

**网上学习资料一大堆，但如果学到的知识不成体系，遇到问题时只是浅尝辄止，不再深入研究，那么很难做到真正的技术提升。**

**[需要这份系统化的资料的朋友，可以添加戳这里获取](https://bbs.csdn.net/topics/618668825)**


**一个人可以走的很快，但一群人才能走的更远！不论你是正从事IT行业的老鸟或是对IT行业感兴趣的新人，都欢迎加入我们的的圈子（技术交流、学习资源、职场吐槽、大厂内推、面试辅导），让我们一起学习成长！**

1715737718562)]
[外链图片转存中...(img-EbhUTRlK-1715737718563)]

**网上学习资料一大堆，但如果学到的知识不成体系，遇到问题时只是浅尝辄止，不再深入研究，那么很难做到真正的技术提升。**

**[需要这份系统化的资料的朋友，可以添加戳这里获取](https://bbs.csdn.net/topics/618668825)**


**一个人可以走的很快，但一群人才能走的更远！不论你是正从事IT行业的老鸟或是对IT行业感兴趣的新人，都欢迎加入我们的的圈子（技术交流、学习资源、职场吐槽、大厂内推、面试辅导），让我们一起学习成长！**