ROS24visard

Z_LASER_PROJECTOR: a package for a laser projector to support with augmented reality regular industrial applications
March 9, 2020
ROS2HMI
March 9, 2020

3D perception using rc_visard

Roboception offers innovative navigation, real-time perception and manipulation solutions for robotic systems, thanks to a combination of hardware and software products. Our sensor, rc_visard (http://roboception.com/en/rc_visard-en/), provides ego-motion, depth data and point clouds in real time. Thanks to fusion of inertial and stereo odometry, a robust pose estimation is possible at all times. Moreover, our product inherits a confidence value with each measurement. This value enables decision making in robotic systems, as it shows how much the sensor reading can be trusted. The sensor is IP54 certified, therefore it is well suited for application in industrial automation and direct integration into the real-time control loop in tasks such as robot navigation and mobile manipulation.

The rc_visard has an existing ROS interface developed within VISARD4ROS, a ROSIN FTP (http://rosin-project.eu/ftp/visard4ros). During this FTP, different components were developed, including the ROS node to connect to the sensor, retrieve the generated information and visualizing it in RVIZ. Several tutorials were also created and posted on ROS wiki (https://wiki.ros.org/rc_visard/Tutorials), and a video that demonstrates the integration of rc_visard with ROS was posted on Youtube (https://www.youtube.com/watch?v=wdK23-gapqw). A corresponding software framework that exploits the capabilities of the sensor, hereafter generically called rc_reason, was developed in 2018. This suite provides software modules that enhance the applicability of the rc_visard to different fields, including TagDetect (detection and pose estimation of QR codes and AprilTags), ItemPick (model-free computation of grasping points for suction cup grasping in bin picking scenarios), BoxPick (detection of grasping points for boxes in palletizing/depalletizing scenarios) and SilhouetteMatch (detection of objects lying on a flat surface). All major processing steps are performed on the on-board processing unit. ROS Interfaces for these modules were developed within another FTP project called reason4ROS.

The objective of ROS24visard is to develop the ROS2 connectivity for expert users and robot developers to fully exploit the sensor and software capabilities of the rc_visard sensor and integrate it into their prototypes or research platforms.