eduzhai > Applied Sciences > Engineering >

Generalized Grasping for Mechanical Grippers for Unknown Objects with Partial Point Cloud Representations

  • Save

... pages left unread,continue reading

Document pages: 8 pages

Abstract: We present a generalized grasping algorithm that uses point clouds (i.e. agroup of points and their respective surface normals) to discover grasp posesolutions for multiple grasp types, executed by a mechanical gripper, in nearreal-time. The algorithm introduces two ideas: 1) a histogram of finger contactnormals is used to represent a grasp shape to guide a gripper orientationsearch in a histogram of object(s) surface normals, and 2) voxel gridrepresentations of gripper and object(s) are cross-correlated to match fingercontact points, i.e. grasp size , to discover a grasp pose. Constraints, suchas collisions with neighbouring objects, are optionally incorporated in thecross-correlation computation. We show via simulations and experiments that 1)grasp poses for three grasp types can be found in near real-time, 2) grasp posesolutions are consistent with respect to voxel resolution changes for bothpartial and complete point cloud scans, and 3) a planned grasp is executed witha mechanical gripper.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...