This package can segment object clusters on top of a table plane.

== ROS API ==
{{{
#!clearsilver CS/NodeAPI
 name = table_object_cluster
 desc = The `table_object_cluster` node takes in <<MsgLink(sensor_msgs/PointCloud2)>> messages for the point map and the table plane and extracts objects clusters on top of the table. As output, bounding boxes for all the objects are provided.

goal{
 0.name=table_object_cluster/goal
 0.type=cob_3d_mapping_msgs/TableObjectClusterActionGoal
 0.desc=contains the table plane and triggers cluster detection
}

result{
 0.name=table_object_cluster/result
 0.type=cob_3d_mapping_msgs/TableObjectClusterActionResult
 0.desc=returns the bounding boxes for the object clusters
}

feedback{
 0.name=table_object_cluster/feedback
 0.type=cob_3d_mapping_msgs/TableObjectClusterActionFeedback
 0.desc=empty
}

sub{
 0.name = /point_cloud
 0.type = sensor_msgs/PointCloud2
 0.desc = The input point cloud
 1.name = /shape_array
 1.type = cob_3d_mapping_msgs/ShapeArray
 1.desc = The supporting plane(s)
}

pub{
 0.name = /bb_marker
 0.type = visualization_msgs/Marker
 0.desc = Bounding boxes for object cluster
 1.name = /cluster_array
 1.type = cob_perception_msgs/PointCloud2Array
 1.desc = The object clusters
}

param{
 0.name = ~height_min
 0.type = double
 0.default = -0.5
 0.desc = lower boundary of cluster search area (relative to table plane, negative values for area on top of table)
 1.name = ~height_max
 1.type = double
 1.default = -0.03
 1.desc = upper boundary of cluster search area (relative to table plane, negative values for area on top of table)
 2.name = ~min_cluster_size
 2.type = double
 2.default = 10
 2.desc = minimum number of points for a valid cluster
 3.name = ~cluster_tolerance
 3.type = double
 3.default = 0.06
 3.desc = threshold for separation of two clusters
 4.name = ~save
 4.type = bool
 4.default = false
 4.desc = Flag for saving the results as PCD files (for debugging)
 5.name = ~file_path
 5.type = string
 5.default = /tmp
 5.desc = File path for saving the results
}

req_tf{
}

prov_tf{
}

}}}

== Usage/Examples ==
=== Action mode and geometry map ===
Start the mapping pipeline in [[cob_3d_mapping_pipeline]]
{{{
roslaunch cob_3d_mapping_pipeline mapping.launch
}}}
Make sure that a table is in the field of view of the camera and trigger the mapping by
{{{
rosrun cob_3d_mapping_point_map trigger_mapping.py start
}}}
Check in rviz if the table is in the map (as a plane).<<BR>>
Start the semantic extraction of tables
{{{
roslaunch cob_3d_mapping_semantics extract_semantics.launch
}}}
Launch the cluster detection
{{{
roslaunch cob_table_object_cluster tabletop_object_cluster.launch
}}}
Run
{{{
rosrun cob_table_object_cluster table_object_cluster_action_client
}}}
This will extract object cluster on each table surface found. Bounding boxes for the cluster are published as <<MsgLink(visualization_msgs/Marker)>>.

=== Topic mode ===
Run the segmentation from [[cob_3d_segmentation]]
{{{
roslaunch cob_3d_segmentation fast_segmentation.launch
}}}
Make sure to subscribe directly to the camera point cloud topic if you are not using registration.
Launch the supporting plane extraction in [[cob_3d_mapping_semantics]]
{{{
roslaunch cob_3d_mapping_semantics extract_supporting_plane.launch
}}}
Launch the object cluster detection
{{{
roslaunch cob_table_object_cluster tabletop_object_cluster.launch
}}}
The object clusters are published as <<MsgLink(visualization_msgs/Marker)>> for bounding boxes and <<MsgLink(cob_perception_msgs/PointCloud2Array)>>.