Research on Object Grasping Point Selection Based on Deep Learning

The research on robot grasping involves mechanical, control, computer, artificial intelligence and so on. Robot Grasping is also a good emplementation of minimal research to support other related research. Efforts on flexibility and interactivity of robot grasping can promote many related studies. In this paper, convolutional neural network is used to study the grasp candidates selection of two-finger robot grasp. The experient explained the processes of select candidate points in detail, the results of the convolutional neural network in the grasp candidates selection is verified through experiments. Keywords—deep learning; depth image; grasping points

• Publication year

  2018

• Venue

  Unknown venue

• Publication date

  2018-03-01

• Fields of study

  Computer Science, Engineering

• Identifiers
  DOI 10.2991/ICAITA-18.2018.16
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Dex-Net 2.0: Deep Learning to Plan Robust Grasps with Synthetic Point Clouds and Analytic Grasp Metrics

  2017cited by this paper
• Shape completion enabled robotic grasping

  2016cited by this paper
• Using Geometry to Detect Grasp Poses in 3D Point Clouds

  2015influential reference
• Learning of grasp selection based on shape-templates

  2013cited by this paper
• Learning a dictionary of prototypical grasp-predicting parts from grasping experience

  2013cited by this paper
• An SVM learning approach to robotic grasping

  2004cited by this paper
• Part pose statistics: estimators and experiments

  1999cited by this paper
• Finding antipodal point grasps on irregularly shaped objects

  1992cited by this paper

• No citing papers are available for this paper.