Real Time Eye Tracking and Hand Tracking Using Regular Video Cameras for Human Computer Interaction

Abstract

This Final Technical Report discusses the accomplishments of a research effort to develop a system for real time eye tracking and hand pointing tracking using regular cameras for human computer interaction. Several novel algorithms for eye detection and eyeball estimation were implemented, including a new model for eye gaze estimation and eye detection. The system allows for viewing direction estimation and eye-based control of screen activity. A novel hand pointing estimation and hand feature detection approach has also been developed. A new approach for hand region detection was designed using Haar-like features, and allow the hand tracking in two views from two cameras. The detailed hand features are also detected and tracked automatically. The developed systems can run in real time.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 2011
Accession Number
ADA536812

Entities

People

  • Lijun Yin

Organizations

  • Binghamton University

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Air Force
  • Air Force Research Laboratories
  • Algorithms
  • Cameras
  • Computational Science
  • Computer Science
  • Computer Vision
  • Detection
  • Human-Computer Interaction
  • Human-Machine Interaction
  • Human-Robot Interaction
  • Information Systems
  • Machine Learning
  • Pattern Recognition
  • Probability
  • Three Dimensional
  • Video Cameras

Fields of Study

  • Computer science

Readers

  • Brain and Cognitive Science; Experimental Psychology; Cognitive Neuroscience
  • Computer Vision.