Location:Home > Engineering science > Electrical >  Power Electronics and Power Drives > The Eye-mouse Research and Implementation Based on Eye Gaze Tracking Technology
Details
Name

The Eye-mouse Research and Implementation Based on Eye Gaze Tracking Technology

Downloads: []
Author
Tutor: DongXiuCheng
School: West China University
Course: Power Electronics and Power Drives
Keywords: Gaze tracking,Human-Computer Interaction,Support Vector Machine,Iris
CLC: TP391.41
Type: Master's thesis
Year:  2011
Facebook Google+ Email Gmail Evernote LinkedIn Twitter Addthis

not access Image Error Other errors

Abstract:
Computer vision is a science, further study how to make machines \With the popularity of computers and electronic products, and its performance continues to improve, more and more people are committed to research in the field of computer vision. Based on the detection of line of sight of the human eye tracking technology has been widely applied to the field of driver fatigue detection, people with disabilities live interaction and entertainment, visual games. This article describes the principles and methods of the video image sequence human eye gaze tracking, dynamic face tracking, eye fixation point calibration made in-depth study, more suitable for video images of the human eye sight tracking algorithm, and combined OpenCV open source code VC 6.0 as a platform to establish a line of sight as the input signal of simple human-computer interaction systems. This article is for the detection of the human face and the human eye, machine learning method based on support vector machine. In order to improve the speed of face tracking the introduction of color model-based adaptive mean shift (Camshift) tracking algorithm on subsequent frames the face tracking Camshift algorithm susceptible to skin color effects scene, the paper presents the improved algorithm, the Camshift algorithm template matching to eliminate the interference of the class color area purposes. Experimental results show that the improved algorithm for real-time and anti-jamming capability is better than the original algorithm, human faces in complex scenes is accurate, fast-track. Eye location and tracking based on the positioning of the image of the iris of the eye circle detection method based on the Hough transform (Hough), and get to the coordinates of the center of the iris in the video window coordinates. Mapping coordinates to the line of sight between the fixation point calibration for the center of the iris, the article proposes a calibration method based on support vector regression line of sight, abandon the traditional polynomial calibration, to solve the mapping relationship that may exist in the small sample fitting and owe fitting problems and improve the accuracy of the sight line calibration. Development of human-computer interaction system based on the methods described herein, to achieve real-time human eye gaze tracking, and human-computer interaction detected the sight. The system is not excessive demands on the hardware, to adapt the illumination changes and head a slight rotation and pitch change, and has good robustness. The system compared with the the mainstream gaze tracking system it has the advantage of low-cost high-efficiency characteristics.
Related Dissertations
Last updated
Sponsored Links
Home |About Us| Contact Us| Feedback| Privacy | copyright | Back to top