Non-Invasive Techniques for Monitoring Human Fatigue

Abstract

In this report, we summarize our efforts in developing real time non-intrusive technology-for monitoring human fatigue. Through this research, we have developed state of the art technologies and a prototype fatigue monitor for real time non-intrusive human fatigue monitoring. Our contributions include: 1) the development of various computer vision techniques for real-time and non-intrusive extraction of multiple fatigue parameters related to eyelid movements, gaze, head movement, and facial expressions, 2) the development of a probabilistic framework based on the Bayesian networks to model and integrate contextual and visual cues information for robust and accurate fatigue detection, and 3) systematic and scientific validation of the fatigue monitor. Experimental validation of our techniques using human subjects demonstrates the good measurement accuracy of our techniques. In addition, the validation also verifies the validity of the proposed fatigue parameters as well as that of the composite fatigue index computed by our fatigue monitor.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 2003
Accession Number
ADA422007

Entities

People

  • Qiang Ji

Organizations

  • University of Nevada, Reno

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Artificial Intelligence
  • Asthenopia
  • Bayesian Networks
  • Composite Materials
  • Computer Science
  • Computer Vision
  • Computers
  • Detection
  • Extraction
  • Heart Rate
  • Human-Machine Interaction
  • Measurement
  • Models
  • Monitoring
  • Probability

Fields of Study

  • Computer science

Readers

  • Circadian Sleep-Wake Regulation and Chronobiology
  • Computational Modeling and Simulation
  • Sensor Fusion and Tracking Systems.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference