Home /Research /Command Recognition of Robots Based on Feature Quantity Extraction from Haptic Information
HRI

Command Recognition of Robots Based on Feature Quantity Extraction from Haptic Information

Toshiaki Tsuji, Tatsuya Itô

Year
2010
Citations
4
Access
Open access

Abstract

In view of safety issues, haptic sensing is inevitable for human support robots. Furthermore, haptic communication based on haptic sensing technology has high potential as a part of multi-modal communication. This paper therefore proposes a method of command recognition based on haptic interfaces. When an operator runs a finger across the haptic interface, the command is recognized based on some feature quantities derived from the contact trajectory. The command recognition method accomplishes a reliable emergency stop function by utilizing force information. Some experimental results show that the efficiency of robot operation is enhanced since multi-dimensional information can be transmitted by a single intuitive motion.

Keywords

Haptic technologyRobotComputer scienceFeature (linguistics)TrajectoryArtificial intelligenceInterface (matter)Feature extractionComputer visionMotion (physics)

Related papers

Browse all HRI papers