Example demonstrating the workflow for grasp planning on a phyiscal robot using point cloud registration. Indexes the most robust grasp from the database and transforms it into the robot frame of reference.


Alan Li and Jeff Mahler

YAML Configuration File Parameters

database : str
full path to a Dex-Net HDF5 database
dataset : str
name of the dataset containing the object instance to grasp
gripper : str
name of the gripper to use
metric : str
name of the grasp robustness metric to use
object : str
name of the object to use (in practice instance recognition is necessary to determine the object instance from images)