The hand shape recognition of human computer interaction with artificial neural network

  • Authors:

  • Jinwen Wei;Hequn Qin;Junjie Guo;Jinwen Wei;Yanling Chen

  • Affiliations:

  • State Key Laboratory for Manufacturing Systems Engineering, Xi’an JiaoTong University, Xi’an, China;State Key Laboratory for Manufacturing Systems Engineering, Xi’an JiaoTong University, Xi’an, China;State Key Laboratory for Manufacturing Systems Engineering, Xi’an JiaoTong University, Xi’an, China;School of Mechanical Engineering, Guangxi University, Nanning, China;School of Mechanical Engineering, Guangxi University, Nanning, China

  • Venue:
  • VECIMS'09 Proceedings of the 2009 IEEE international conference on Virtual Environments, Human-Computer Interfaces and Measurement Systems
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

The hand gestures used in Human Computer Interaction (HCI) are generally posed by complicated and large amplitude actions of arm /hand. Thus usable HCI instructions are few and HCI efficiency is low. This paper presents new hand shapes and the corresponding recognition system for the HCI with robot or Coordinate Measuring Machine. Using a touchpad to precept the touching of fingers, hand shapes posed to express HCI instructions are defined by the combinations of 2 binary status, i.e. status of touching /detaching on touchpad and status of stretching /retracting over touchpad, of Index, Middle, Ring and Little fingers. Method of extracting the features in hand shape image is presented. Based on Neural Network, a decision binary tree is used in the real-time recognition of the hand shapes. A correctness ratio of about 95% is obtained when implemented by DSP processor in the recognition of 12 hand shapes.