【踩坑记录】仿真环境使用小车进行Cartographer 3D Slam(深度摄像头)
所有内容请看:
【运行背景】
ROS1 20.04 noetic
安装cartographer请看:
【安装学习】安装Cartographer ROS(noetic)_Howe_xixi的博客-CSDN博客网上使用noetic安装cartographer的比较少,所以将安装步骤记录下来
https://blog.csdn.net/weixin_44362628/article/details/122540297?spm=1001.2014.3001.5501gazebo仿真环境:如有以下问题请看:
目录
【Cartographer配置】
【修改lua文件】
cartographer_ros配置文件.lua文件路径位于cartographer_ros/configuration_files中,选择其中的backpack_3d.lua文件复制修改,取新名字为spark_pointcloud.lua,内容如下:
-- Copyright 2016 The Cartographer Authors
--
-- Licensed under the Apache License, Version 2.0 (the "License");
-- you may not use this file except in compliance with the License.
-- You may obtain a copy of the License at
--
-- http://www.apache.org/licenses/LICENSE-2.0
--
-- Unless required by applicable law or agreed to in writing, software
-- distributed under the License is distributed on an "AS IS" BASIS,
-- WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-- See the License for the specific language governing permissions and
-- limitations under the License.
include "map_builder.lua"
include "trajectory_builder.lua"
options = {
map_builder = MAP_BUILDER,
trajectory_builder = TRAJECTORY_BUILDER,
map_frame = "map",
tracking_frame = "base_footprint",
published_frame = "odom",
odom_frame = "odom",
provide_odom_frame = false,
publish_frame_projected_to_2d = false,
use_pose_extrapolator = 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 = 5,
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.,
fixed_frame_pose_sampling_ratio = 1.,
imu_sampling_ratio = 1.,
landmarks_sampling_ratio = 1.,
}
TRAJECTORY_BUILDER_3D.num_accumulated_range_data = 1
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
主要修改以下内容:
a) racking_frame = "base_footprint",仿真时用base_footprint,实体时用imu_link;
b) TRAJECTORY_BUILDER_3D.num_accumulated_range_data = 1,设置太大会导致数据处理错误
*其他参数值含义可参考官网,链接见以下链接
【修改launch文件】
cartographer_ros的launch文件backpack_3d.launch,路径位于/cartographer_ros/launch中,选择并进行复制修改,取新名字spark_cartographer_3d.launch,内容如下:
<?xml version="1.0"?>
<launch>
<node name="cartographer_node" pkg="cartographer_ros"
type="cartographer_node" args="
-configuration_directory $(find cartographer_ros)/configuration_files
-configuration_basename spark_pointcloud.lua"
output="screen">
<remap from="/points2" to="/camera/depth/points" />
</node>
<node name="cartographer_occupancy_grid_node" pkg="cartographer_ros"
type="cartographer_occupancy_grid_node" args="-resolution 0.05" />
<node name="rviz" pkg="rviz" type="rviz" required="true"
args="-d $(find cartographer_ros)/configuration_files/demo_3d.rviz" />
</launch>
主要启动以下内容:
a) 启动cartographer_node节点,配置文件指向刚刚设置的.lua文件,同时映射points2对应的话题
b) 启动cartographer_occupancy_grid_node地图格式转换节点,由于cartographer_node建图节点提供的地图是submapList格式的,需要转换成GridMap格式才能在ROS中显示和使用。这里面有两个可配参数,resolution用来设置GridMap地图的分辨率,publish_period_sec用来设置GridMap地图发布的频率。
c) 启动Rviz,启动成功后,再根据需求保存配置即可
【重新编译】
通过Lua脚本配置参数的方法,每次修改参数后需要重新编译,否则参数无法生效
catkin_make_isolated --install --use-ninja
【Gazebo仿真】
均在spark-slam-gazebo工作空间下
【小车URDF模型】
Cartographer需要imu和pointcloud的数据,所以添加这两部分的传感器
a)添加深度摄像头
<link name="virtual_camera_link" />
<joint name="virtual_camera_joint"
type="fixed">
<origin
xyz="0 0 0.03"
rpy="0 0 0" />
<parent link="camera_Link" />
<child link="virtual_camera_link" />
</joint>
<gazebo reference="virtual_camera_link">
<sensor type="depth" name="camera">
<always_on>true</always_on>
<update_rate>20.0</update_rate>
<camera>
<horizontal_fov>${60.0*3.14/180.0}</horizontal_fov>
<image>
<format>R8G8B8</format>
<width>640</width>
<height>480</height>
</image>
<clip>
<near>0.05</near>
<far>5.0</far>
</clip>
</camera>
<plugin name="kinect_camera_controller" filename="libgazebo_ros_openni_kinect.so">
<cameraName>camera</cameraName>
<alwaysOn>true</alwaysOn>
<updateRate>20.0</updateRate>
<imageTopicName>rgb/image_raw</imageTopicName>
<depthImageTopicName>depth/image_raw</depthImageTopicName>
<pointCloudTopicName>depth/points</pointCloudTopicName>
<cameraInfoTopicName>rgb/camera_info</cameraInfoTopicName>
<depthImageCameraInfoTopicName>depth/camera_info</depthImageCameraInfoTopicName>
<frameName>upper_kinect_pointcloud_link</frameName>
<baseline>0.1</baseline>
<distortion_k1>0.0</distortion_k1>
<distortion_k2>0.0</distortion_k2>
<distortion_k3>0.0</distortion_k3>
<distortion_t1>0.0</distortion_t1>
<distortion_t2>0.0</distortion_t2>
<pointCloudCutoff>0.4</pointCloudCutoff>
</plugin>
</sensor>
</gazebo>
b) 添加imu
<link name="imu_link">
<visual>
<origin xyz="0.0 0.0 0.0"/>
<geometry>
<cylinder length="0.07" radius="0.05"/>
</geometry>
</visual>
</link>
<joint
name="imu_joint"
type="fixed">
<origin
xyz="-0.0 0.0 0.0"
rpy="0 0 0" />
<parent
link="base_footprint" />
<child
link="imu_link" />
<axis
xyz="0 1 0" />
</joint>
<gazebo reference="imu_link">
<gravity>true</gravity>
<sensor name="imu_sensor" type="imu">
<always_on>true</always_on>
<update_rate>20</update_rate>
<visualize>true</visualize>
<topic>__default_topic__</topic>
<plugin filename="libgazebo_ros_imu_sensor.so" name="imu_plugin">
<topicName>imu</topicName>
<bodyName>imu_link</bodyName>
<updateRateHZ>20</updateRateHZ>
<gaussianNoise>0.0</gaussianNoise>
<rpyOffset>0 0 0</rpyOffset>
<frameName>imu_link</frameName>
</plugin>
<pose>0 0 0 0 0 0</pose>
</sensor>
</gazebo>
【编写launch文件】
取名为spark_cartographer_slam.launch,内容如下:
<?xml version="1.0"?>
<launch>
<!-- 设置launch文件的参数 -->
<arg name="paused" default="false"/>
<arg name="use_sim_time" default="true"/>
<arg name="gui" default="true"/>
<arg name="headless" default="true"/>
<arg name="debug" default="false"/>
<arg name="model" default="$(find spark1_description)/urdf/spark1_description_cartographer.urdf"/>
<!-- 运行gazebo仿真环境 -->
<include file="$(find gazebo_ros)/launch/empty_world.launch">
<arg name="debug" value="$(arg debug)" />
<arg name="gui" value="$(arg gui)" />
<arg name="paused" value="$(arg paused)"/>
<arg name="use_sim_time" value="$(arg use_sim_time)"/>
<arg name="headless" value="$(arg headless)"/>
</include>
<!-- 运行joint_state_publisher节点,发布机器人的关节状态 -->
<node name="joint_state_publisher" pkg="joint_state_publisher" type="joint_state_publisher" ></node>
<!-- 运行robot_state_publisher节点,发布tf -->
<node name="robot_state_publisher" pkg="robot_state_publisher" type="robot_state_publisher" output="screen" >
<param name="publish_frequency" type="double" value="50.0" />
</node>
<!-- 在gazebo中加载机器人模型-->
<param name="robot_description" command="$(find xacro)/xacro --inorder $(arg model)" />
<node name="urdf_spawner" pkg="gazebo_ros" type="spawn_model" respawn="false" output="screen"
args="-unpause -urdf -model marobot1 -param robot_description -x 0.0 -y 0.0"/>
<!--创建新的终端,spark键盘控制 “wsad”分别代表“前后左右”-->
<node pkg="spark_teleop" type="keyboard_control.sh" name="kc_2d" />
</launch>
该launch文件包含以下五个部分:
- 运行Gazebo仿真环境
- 运行joint_state_publisher节点,发布机器人的关节状态
- 运行robot_state_publisher节点,发布tf
- 在gazebo中加载机器人模型
- 运行键盘控制节点
【运行Cartographer仿真】
a ) 启动仿真环境
roslaunch spark1_description spark_cartographer_slam.launch

添加上自己搭的简易模型,添加障碍物
b ) 启动Cartographer
roslaunch cartographer_ros spark_cartographer_3d.launch

机器人出现漂移现象,导致建图效果不好,后续会进行优化
【问题总结】
【如何添加传感器】
在gazebo中可以通过插入一些插件,来仿真机器人的传感器、执行器的特性,这些插件通过<gazebo>元素中的<plugin>标签描述,使用下边的格式加入到URDF文件中(详细请看上文添加camera和imu):
<gazebo reference="your_link_name">
<gazebo>
<plugin name="kinect_camera_controller" filename="libgazebo_ros_openni_kinect.so">
... plugin parameters ...
</plugin>
</gazebo>
<gazebo reference>指的是插件附着的位置,比如机器人的某个link,gazebo默认支持不少常用的设备,而且也放置在ros安装目录下(也有在gazebo_plugins功能包中)

同时添加一个虚拟的link和joint将插件固定在机器人上
<link name="virtual_camera_link" />
<joint name="virtual_camera_joint"
type="fixed">
<origin
xyz="0 0 0.03"
rpy="0 0 0" />
<parent link="camera_Link" />
<child link="virtual_camera_link" />
</joint>
【解决点云显示错误的问题】
曾经出现过点云显示位置错误,调用camera查看图形没有问题,在rviz上显示错误,如下图:

解决办法:
a)添加一个虚拟的关节,关联到摄像机的link,调整rpy值(rqy分别对应着xyz轴,遵循右手定理),将点云调整到正确位置
<link name="upper_kinect_pointcloud_link"/>
<joint name="upper_kinect_pointcloud_joint" type="fixed">
<origin xyz="0 0 0" rpy="-1.57 0 -1.57"/>
<parent link="virtual_camera_link"/>
<child link="upper_kinect_pointcloud_link"/>
</joint>
b)将深度摄像头插件中的frameName改为新增加的link
<plugin name="kinect_camera_controller" filename="libgazebo_ros_openni_kinect.so">
<cameraName>camera</cameraName>
<alwaysOn>true</alwaysOn>
<updateRate>20.0</updateRate>
<imageTopicName>rgb/image_raw</imageTopicName>
<depthImageTopicName>depth/image_raw</depthImageTopicName>
<pointCloudTopicName>depth/points</pointCloudTopicName>
<cameraInfoTopicName>rgb/camera_info</cameraInfoTopicName>
<depthImageCameraInfoTopicName>depth/camera_info</depthImageCameraInfoTopicName>
<frameName>upper_kinect_pointcloud_link</frameName>
<baseline>0.1</baseline>
<distortion_k1>0.0</distortion_k1>
<distortion_k2>0.0</distortion_k2>
<distortion_k3>0.0</distortion_k3>
<distortion_t1>0.0</distortion_t1>
<distortion_t2>0.0</distortion_t2>
<pointCloudCutoff>0.4</pointCloudCutoff>
</plugin>
再次运行仿真环境,点云就会正确显示
【感谢】
ROS探索总结(二十四)——使用gazebo中的插件 - 古月居
Gazebo urdf kinetic RGBD camera 相机视角和点云视角不一致_Lucky Vivi的博客-CSDN博客
更多推荐
所有评论(0)