pcl_matching

The package is used to implement basic online mapping using PCL library.

Requirements

1. Eigen3
2. PCL

Nodes

1. Node ndt_mapping_node
2. Node icp_mapping_node

1. Node ndt_mapping_node

(a) Setup

Before we start running the mapping module, there is one configuration file needed to be specfied.

1. path:
   Their path is

~/pcl_matching/config/config.yaml

2. format:
   config.yaml

topic_name_lidar: "/points_raw"
map_saved: true
map_resolution: 0.02
map_update:
displacement: 5
time_difference: 1

topic_name_lidar is the topic name of point cloud.
map_saved is used to indicate the node to save the current map.
(i.e. If you intend to save the current map, type rosparam set /ndt_mapping_node/map_saved false, it'll save the current map in ~/pcl_matching/data/map.pcd.)
map_resolution is the resolution of built map. It use voxel grid filter to downsample the map point cloud.
map_update/displacement and map_update/time_difference determines when to update map (add new point cloud to map), either by the displacement or time difference to last updated pose.

(b) Getting Started.

1. Launch the node

roslaunch pcl_matching ndt_mapping.launch

(c) Subscribed Topics

• /points_raw (sensor_msgs/PointCloud2)
  Point cloud for building map.

(d) Published Topics

• /ndt_mapping_node/local_map (sensor_msgs/PointCloud2)
  Built map. The first LiDAR frame is set as the map frame.

2. Node icp_mapping_node

(a) Setup

Before we start running the mapping module, there is one configuration file needed to be specfied.

1. path:
   Their path is

~/pcl_matching/config/config.yaml

2. format:
   config.yaml

topic_name_lidar: "/points_raw"
map_saved: true
map_resolution: 0.02
map_update:
displacement: 5
time_difference: 1

topic_name_lidar is the topic name of point cloud.
map_saved is used to indicate the node to save the current map.
(i.e. If you intend to save the current map, type rosparam set /icp_mapping_node/map_saved false, it'll save the current map in ~/pcl_matching/data/map.pcd.)
map_resolution is the resolution of built map. It use voxel grid filter to downsample the map point cloud.
map_update/displacement and map_update/time_difference determines when to update map (add new point cloud to map), either by the displacement or time difference to last updated pose.

(b) Getting Started.

1. Launch the node

roslaunch pcl_matching icp_mapping.launch

(c) Subscribed Topics

• /points_raw (sensor_msgs/PointCloud2)
  Point cloud for building map.

(d) Published Topics

• /icp_mapping_node/local_map (sensor_msgs/PointCloud2)
  Built map. The first LiDAR frame is set as the map frame.

• Result of ndt mapping