VoxelGridDownsampleManager

What Is This

../../_images/voxel_grid_downsample.png

Filter input point cloud by input marker size, and then downsample it.

Subscribing Topic

  • ~input (sensor_msgs/PointCloud2)

    Original point cloud.

  • ~add_grid (visualization_msgs/Marker)

    Additional marker used for xyz pass through filter before downsampling.

    Only frame_id, pose/position and scale field are used for filtering.

    Also, leaf size [m] in voxel grid downsampling will be defined by color/r field.

    Below marker is added internally by default.

    header:
      frame_id: ~base_frame
    pose:
      position:
        x: 2.0
        y: 0.0
        z: -0.5
    scale:
      x: 4.0
      y: 2.0
      z: 3.0
    color:
      r: 0.05
    

Publishing Topic

  • ~output (sensor_msgs/PointCloud2)

    Downsampled point cloud for debugging.

  • ~output_encoded (jsk_recognition_msgs/SlicedPointCloud)

    Downsampled and sliced point cloud.

    All sliced clusters will be published in order with the slice_index info.

    Number of the clusters is calculated from ~max_points.

    Messages of this topic can be decoded by jsk_pcl_ros/VoxelGridDownsampleDecoder.

Parameter

  • ~base_frame (String, default: pelvis)

    Frame ID of initial marker.

  • ~max_points (Int, default: 300)

    Number of maximum points in ~output_encoded.

  • ~rate (Float, default: 1.0)

    Multiplicative inverse of duration between publishing output_encoded.

    The unit is [Hz].

Sample

roslaunch jsk_pcl_ros sample_voxel_grid_downsample.launch