A robust and friendly humanrobot interface system based on natural human gestures

Ming Shaung Chang, Jung-Hua Chou

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)

Abstract

In this paper, we design a robust and friendly humanrobot interface (HRI) system for our intelligent mobile robot based only on natural human gestures. It consists of a triple-face detection method and a fuzzy logic controller (FLC)-Kalman filter tracking system to check the users and predict their current position in a dynamic and cluttered working environment. In addition, through the combined classifier of the principal component analysis (PCA) and back-propagation artificial neural network (BPANN), single and successive commands defined by facial positions and hand gestures are identified for real-time command recognition after dynamic programming (DP). Therefore, the users can instruct this HRI system to make member recognition or expression recognition corresponding to their gesture commands, respectively based on the linear discriminant analysis (LDA) and BPANN. The experimental results prove that the proposed HRI system perform accurately in real-time face detection and tracking, and robustly react to the corresponding gesture commands at eight frames per second (fps).

Original languageEnglish
Pages (from-to)847-866
Number of pages20
JournalInternational Journal of Pattern Recognition and Artificial Intelligence
Volume24
Issue number6
DOIs
Publication statusPublished - 2010 Sep 1

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Software
  • Computer Vision and Pattern Recognition

Fingerprint Dive into the research topics of 'A robust and friendly humanrobot interface system based on natural human gestures'. Together they form a unique fingerprint.

Cite this