[Documentation] [TitleIndex] [WordIndex

ais_bonn_pcl_addons: octreelib

Package Summary

The octreelib contains an octree implementation that is templated for coordinate and value type. It features fast sampling and value aggregation on many resolutions in parallel. The implementation is specifically suited for fast volume queries of integral or extremum values.

Properties

largescale.png smallscale.png

Getting started

The package contains a launch file that computes multiresolution normals in a nodelet.

$ roslaunch octreelib normalestimation_nodelet.launch

The input PointCloud2 topic of the nodelet is mapped to /camera/depth/points for Kinect depth streams. The launch file also starts RViz with a proper configuration to visualize output normals from the PointCloud2 output messages and visualization markers.

ROS API

normalestimation_nodelet

normalestimation_nodelet calculates multiresolution normals on arbitrary point clouds.

Subscribed Topics

~input (sensor_msgs/PointCloud2)

Published Topics

~output_layerN (sensor_msgs/PointCloud2) ~visualization_marker_array (visualization_msgs/MarkerArray)

Parameters

~fixed_size (bool, default: true) ~min_resolution (double, default: 0.005) ~max_range (double, default: 30.0) ~min_points_for_fit (int, default: 25)


2011-11-19 12:28