I am trying to add perception capabilities to my `MoveIt!` configuration.
I am running `ros` `indigo`, and I currently have `Gazebo` running, with `MoveIt!` successfully talking to the controllers and moving the (simulated in `Gazebo`) robot.
In order to implement arm movements that do not collide with the environment, I added a camera (`libgazebo_ros_openni_kinect.so`) to my robot description, and followed the [instructions](http://docs.ros.org/indigo/api/pr2_moveit_tutorials/html/planning/src/doc/perception_configuration.html) from the `MoveIt!` tutorials.
Specifically, I added `sensors_rgbd.yaml` to `config` folder:
sensors:
- sensor_plugin: occupancy_map_monitor/PointCloudOctomapUpdater
point_cloud_topic: /camera/camera/depth_registered/points
max_range: 5.0
point_subsample: 1
padding_offset: 0.01
padding_scale: 1.0
filtered_cloud_topic: filtered_cloud
which is successfully loaded via the `sensor_manager.launch.xml` file (I can see the parameters loaded with `rosparam`). Moreover I have the following parameters set:
where `ra_base_link` is a fixed frame in my robot.
I confirmed that the `/camera/camera/depth_registered/points` is indeed published correctly, but I get nothing (for the collision environment) neither in `move_group/monitored_planning_scene` nor in the `RViZ` environment.
Any suggestions?
Cheers,
Nikitas
↧