Abstract—Object grasping of humans and robots is a research topic that contains many challenges. This process needs to solve many problems in the field of computer vision and robotics such as kinetics, hand recognition and positioning, object recognition and positioning in the environment, generate hand shape to grasp, detect grasp area. To perform an effective grasp, it is necessary to understand the attributes of the grasp object. From this taking manipulation actions to fit with the object attributes. In order to make automatic grasp, the steps in (grasp types recognition, object attributes recognition, manipulation actions recognition) need to automatically understand and follow a consistent model. In the paper, we conduct a survey and systematized approaches to solve each component problem based on the relationship of interaction between issues: grasp types recognition, object attributes recognition, manipulation actions recognition. Approaches to solve each problem are presented from traditional methods to modern methods. For example, the training of identification models is presented based on traditional methods such as using SVM to train on characteristics to use deep learning models with the Convolutional Neural Networks (CNNs) for training identification model.
 

Index Terms—grasp types, object attributes, object recognition, relationship

Cite: Van-Hung Le, "Relationship between Grasping Actions and Object Attributes: A Survey," Journal of Advances in Information Technology, Vol. 12, No. 1, pp. 6-13, February 2021. doi: 10.12720/jait.12.1.6-13

Copyright © 2021 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.