Re: 2D LiDAR on quadrotor

Posted by heshamhendy on
URL: http://official-rtab-map-forum.206.s1.nabble.com/2D-LiDAR-on-quadrotor-tp9124p9169.html

the main params are ¸

 <!-- RGB-D related topics -->
  <arg name="rgb_topic"               default="/zed2i/zed_node/rgb/image_rect_color" />
  <arg name="depth_topic"             default="/zed2i/zed_node/depth/depth_registered" />
  <arg name="camera_info_topic"       default="/zed2i/zed_node/depth/camera_info" />
  <arg name="depth_camera_info_topic" default="$(arg camera_info_topic)" />
  
  <!-- stereo related topics -->
  <arg name="stereo_namespace"        default="/stereo_camera"/>
  <arg name="left_image_topic"        default="$(arg stereo_namespace)/left/image_rect_color" />
  <arg name="right_image_topic"       default="$(arg stereo_namespace)/right/image_rect" />      <!-- using grayscale image for efficiency -->
  <arg name="left_camera_info_topic"  default="$(arg stereo_namespace)/left/camera_info" />
  <arg name="right_camera_info_topic" default="$(arg stereo_namespace)/right/camera_info" />

  
  <!-- Already synchronized RGB-D related topic, with rtabmap_ros/rgbd_sync nodelet -->
  <arg name="rgbd_sync"               default="true"/>         <!-- pre-sync rgb_topic, depth_topic, camera_info_topic, set to true if an odom from VIO will be used.  -->
  <arg name="approx_rgbd_sync"        default="false"/>          <!-- false=exact synchronization -->
  <arg name="subscribe_rgbd"          default="$(arg rgbd_sync)"/>
  <arg name="rgbd_topic"              default="rgbd_image" />
  <arg name="depth_scale"             default="1.0" />         <!-- Deprecated, use rgbd_depth_scale instead -->
  <arg name="rgbd_depth_scale"        default="$(arg depth_scale)" />
  <arg name="rgbd_decimation"         default="1" />
  
  <arg name="compressed"              default="false"/>         <!-- If you want to subscribe to compressed image topics -->
  <arg name="rgb_image_transport"     default="compressed"/>    <!-- Common types: compressed, theora (see "rosrun image_transport list_transports") -->
  <arg name="depth_image_transport"   default="compressedDepth"/>  <!-- Depth compatible types: compressedDepth (see "rosrun image_transport list_transports") -->
  
  <arg name="gen_cloud"               default="false"/> <!-- only works with depth image and if not subscribing to scan_cloud topic-->
  <arg name="gen_cloud_decimation"    default="4"/>
  <arg name="gen_cloud_voxel"         default="0.05"/>
   
  <arg name="subscribe_scan"          default="true"/>
  <arg name="scan_topic"              default="/scan"/>
  <arg name="subscribe_scan_cloud"    default="$(arg gen_cloud)"/>
  <arg name="scan_cloud_topic"        default="/scan_cloud"/>
  <arg name="subscribe_scan_descriptor" default="false"/>
  <arg name="scan_descriptor_topic"   default="/scan_descriptor"/>
  <arg name="scan_cloud_max_points"   default="0"/>
  <arg name="scan_cloud_filtered"     default="false"/> <!-- use filtered cloud from icp_odometry for mapping -->
  <arg name="gen_scan"                default="false"/> <!-- only works with depth image and if not subscribing to scan topic-->
  
  <arg name="gen_depth"                  default="false" /> <!-- Generate depth image from scan_cloud -->
  <arg name="gen_depth_decimation"       default="1" />
  <arg name="gen_depth_fill_holes_size"  default="0" />
  <arg name="gen_depth_fill_iterations"  default="1" />
  <arg name="gen_depth_fill_holes_error" default="0.1" />
   
  <arg name="visual_odometry"          default="true"/>          <!-- Launch rtabmap visual odometry node -->
  <arg name="icp_odometry"             default="true"/>         <!-- Launch rtabmap icp odometry node -->
  <arg name="odom_topic"               default="odom"/>          <!-- Odometry topic name like /zed2i/zed_node/odom or /odometry/filtered-->
  <arg name="vo_frame_id"              default="$(arg odom_topic)"/> <!-- Visual/Icp odometry frame ID for TF -->
  <arg name="publish_tf_odom"          default="true"/>            <!-- revise in case of fusion-->>
  <arg name="odom_tf_angular_variance" default="0.001"/>         <!-- If TF is used to get odometry, this is the default angular variance -->
  <arg name="odom_tf_linear_variance"  default="0.001"/>         <!-- If TF is used to get odometry, this is the default linear variance -->
  <arg name="odom_args"                default=""/>              <!-- More arguments for odometry (overwr/zed2i/zed_node/imu/datate same parameters in rtabmap_args) -->
  <arg name="odom_sensor_sync"         default="true"/>
  <arg name="odom_guess_frame_id"        default=""/>
  <arg name="odom_guess_min_translation" default="0"/>
  <arg name="odom_guess_min_rotation"    default="0"/>
  <arg name="odom_max_rate"            default="0"/>
  <arg name="odom_expected_rate"       default="0"/>
  <arg name="imu_topic"                default="/gx5/imu/data"/>          <!-- only used with VIO approaches: ex: /mavros/imu/data (use!!) or /zed2i/zed_node/imu/data  or /gx5/nav/filtered_imu/data--> <!-- there is some thong wrong!!!? -->
  <arg name="wait_imu_to_init"         default="false"/>
  <arg name="use_odom_features"        default="false"/>

¸

The main problem is I would like to use 2D-LIDAR on a drone that moves in Z-axis: The map that gets plotted moves upwards as far as moves it up. The assumption from rtab-map is that having 2D LiDAR means we have a motion on x-y plane! Is there a way to fix that.