Welcome to the InfoSci Platform
Direct Perception and Action Decision for Unknown Object Grasping

Direct Perception and Action Decision for Unknown Object Grasping

Hiroyuki Masuta, Tatsuo Motoyoshi, Kei Sawai, Ken'ichi Koyanagi, Toru Oshima, Hun-Ok Lim
Copyright: © 2019 |Pages: 16
ISBN13: 9781522580607|ISBN10: 1522580603|EISBN13: 9781522580614
DOI: 10.4018/978-1-5225-8060-7.ch064
Cite Chapter Cite Chapter

MLA

Masuta, Hiroyuki, et al. "Direct Perception and Action Decision for Unknown Object Grasping." Rapid Automation: Concepts, Methodologies, Tools, and Applications, edited by Information Resources Management Association, IGI Global, 2019, pp. 1372-1387. https://doi.org/10.4018/978-1-5225-8060-7.ch064

APA

Masuta, H., Motoyoshi, T., Sawai, K., Koyanagi, K., Oshima, T., & Lim, H. (2019). Direct Perception and Action Decision for Unknown Object Grasping. In I. Management Association (Ed.), Rapid Automation: Concepts, Methodologies, Tools, and Applications (pp. 1372-1387). IGI Global. https://doi.org/10.4018/978-1-5225-8060-7.ch064

Chicago

Masuta, Hiroyuki, et al. "Direct Perception and Action Decision for Unknown Object Grasping." In Rapid Automation: Concepts, Methodologies, Tools, and Applications, edited by Information Resources Management Association, 1372-1387. Hershey, PA: IGI Global, 2019. https://doi.org/10.4018/978-1-5225-8060-7.ch064

Export Reference

Mendeley
Favorite

Abstract

This paper discusses the direct perception of an unknown object and the action decision to grasp an unknown object using depth sensor for social robots. Conventional methods estimate the accurate physical parameters when a robot wants to grasp an unknown object. Therefore, we propose a perceptual system based on an invariant concept in ecological psychology, which perceives the information relevant to the action of the robot. Firstly, we proposed the plane detection based approach for perceiving an unknown object. In this paper, we propose the sensation of grasping which is expressed by using inertia tensor, and applied with fuzzy inference using the relation between principle moment of inertia. The sensation of grasping encourages the decision for the grasping action directly without inferring from physical value such as size, posture and shape. As experimental results, we show that the sensation of grasping expresses the relative position and posture between the robot and the object, and the embodiment of the robot arm by one parameter. And, we verify the validity of the action decision from the sensation of grasping.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.