ros noetic使用pointcloud_to_laserscan 将2d激光雷达与深度摄像头数据融合

配置文件

src/wpb_home/wpb_home_tutorials/nav_depth/local_costmap_params.yaml

复制代码
local_costmap:
  global_frame: odom
  robot_base_frame: base_footprint
  static_map: false
  rolling_window: true
  width: 3.0
  height: 3.0
  update_frequency: 10.0
  publish_frequency: 10.0
  transform_tolerance: 1.0
  plugins:
    - {name: obstacle_layer, type: "costmap_2d::ObstacleLayer"}
    - {name: inflation_layer, type: "costmap_2d::InflationLayer"}

配置文件

src/wpb_home/wpb_home_tutorials/nav_depth/costmap_common_params.yaml

复制代码
robot_radius: 0.25

obstacle_layer:
  obstacle_range: 3.0
  raytrace_range: 6.0
  observation_sources: base_lidar depth_scan

  base_lidar:
    data_type: LaserScan
    topic: /scan
    marking: true
    clearing: true

  depth_scan:
    data_type: LaserScan
    topic: /depth_scan
    marking: true
    clearing: true

inflation_layer:
  inflation_radius: 0.5

配置文件

src/wpb_home/wpb_home_tutorials/nav_depth/global_costmap_params.yaml

复制代码
global_costmap:
  global_frame: map
  robot_base_frame: base_footprint
  static_map: false
  rolling_window: true
  update_frequency: 1.0
  publish_frequency: 1.0
  transform_tolerance: 1.0

recovery_behaviors:
  - name: 'conservative_reset'
    type: 'clear_costmap_recovery/ClearCostmapRecovery'
  - name: 'rotate_recovery'
    type: 'rotate_recovery/RotateRecovery'
  - name: 'aggressive_reset'
    type: 'clear_costmap_recovery/ClearCostmapRecovery'

conservative_reset:
  reset_distance: 2.0
  layer_names: ["obstacle_layer"]

aggressive_reset:
  reset_distance: 0.0
  layer_names: ["obstacle_layer"]

配置文件

src/wpb_home/wpb_home_tutorials/nav_depth/depth_to_scan.yaml

复制代码
target_frame: base_footprint
transform_tolerance: 0.3

min_height: 0.10
max_height: 2.00

angle_min: -1.57
angle_max:  1.57
angle_increment: 0.005

scan_time: 0.1
range_min: 0.4
range_max: 6.0
use_inf: true

launch文件

复制代码
<launch>

  <!-- 载入 机器人 和 RoboCup@Home 的仿真场景 -->
  <include file="$(find wpr_simulation)/launch/wpb_stage_robocup.launch"/>

  <!-- Gmapping -->
  <node pkg="gmapping" type="slam_gmapping" name="slam_gmapping"/>

  <!-- Rviz -->
  <arg name="rvizconfig" default="$(find wpr_simulation)/rviz/slam.rviz" />
  <node name="rviz" pkg="rviz" type="rviz" args="-d $(arg rvizconfig)" required="true" />

  <!-- 手柄控制 -->
  <node respawn="true" pkg="joy" type="joy_node" name="joy_node" >
    <param name="dev" type="string" value="/dev/input/js0" />
    <param name="deadzone" value="0.12" />
  </node>
  <param name="axis_linear" value="1" type="int"/>
  <param name="axis_angular" value="0" type="int"/>
  <param name="scale_linear" value="0.5" type="double"/>
  <param name="scale_angular" value="1" type="double"/>
  <node pkg="wpr_simulation" type="teleop_js_node" name="teleop_js_node"/>

      <node pkg="move_base" type="move_base" name="move_base">
        <rosparam file="$(find wpb_home_tutorials)/nav_depth/costmap_common_params.yaml" command="load" ns="global_costmap" />
        <rosparam file="$(find wpb_home_tutorials)/nav_depth/costmap_common_params.yaml" command="load" ns="local_costmap" />
        <rosparam file="$(find wpb_home_tutorials)/nav_depth/global_costmap_params.yaml" command="load" />
        <rosparam file="$(find wpb_home_tutorials)/nav_depth/local_costmap_params.yaml" command="load" />
        <param name="base_global_planner" value="global_planner/GlobalPlanner" /> 
        <param name="base_local_planner" value="wpbh_local_planner/WpbhLocalPlanner" />
    </node>

      <node pkg="pointcloud_to_laserscan"
        type="pointcloud_to_laserscan_node"
        name="pointcloud_to_laserscan"             
        output="screen">
    <!-- name="pointcloud_to_laserscan" 将命名空间载入 rosparam,所以-->
    <!-- 载入参数 -->
    <rosparam file="$(find wpb_home_tutorials)/nav_depth/depth_to_scan.yaml" />

    <!-- 输入点云 -->
    <remap from="cloud_in" to="/kinect2/sd/points"/>

    <!-- 输出 LaserScan -->
    <remap from="scan" to="/depth_scan"/>
  </node>

      <node pkg="wpr_simulation"
      type="demo_map.py"
      name="explore_then_return"
      output="screen"
      >
      </node>
      
            <node pkg="wpr_simulation"
      type="save_map_and_start_amcl.py"
      name="save_map_and_start_amcl"
      output="screen"
      >
      </node>
                  <node pkg="wpr_simulation"
      type="initialpose_service.py"
      name="initialpose_service"
      output="screen"
      >
      </node>
</launch>
相关推荐
MIXLLRED4 小时前
Ubuntu22.04 + ROS2 Humble + RealSense D435i 部署VINS-Fusion视觉惯性SLAM
ubuntu·slam·d435i·ros2·humble·vins
AGV算法笔记1 天前
CVPR 2025顶级SLAM论文精读:MASt3R-SLAM如何用单目相机实现实时稠密三维重建?
深度学习·数码相机·机器人视觉·slam·三维重建·agv
大势智慧2 天前
智影R200手持SLAM使用教程八:3DGS数据采集规范
经验分享·教程·数据采集·slam·3dgs·三维扫描·三维激光扫描仪
G果2 天前
SLAM 开源算法汇总(支持ROS2)
学习·算法·slam·ros2
CS_Zero8 天前
无人机避障——MID360+FasterLIO+EGO-planner实测问题解决
slam·无人机避障·lio
Maxwell的猫9 天前
激光雷达:给自动驾驶装上“火眼金睛”的终极传感器
自动驾驶·汽车·激光雷达·辅助驾驶
冰水不凉10 天前
robot_localization实现imu和odom融合
前端·slam
大势智慧12 天前
智影R200手持slam使用教程二:开始与结束采集
教程·数据采集·slam·开机·三维扫描·三维激光扫描仪
胡摩西12 天前
室内定位技术方法汇总:从WiFi到超声波,机器人如何在室内“找准自己”?
人工智能·机器人·slam·室内定位·roomaps
加油JIAX13 天前
为什么边缘化先验的Hessian矩阵可分解得到Jacobian矩阵?
slam·滑动窗口·边缘化