A Study Of Gaze Tracking Technology In Human-Computer Interaction

In recent years, the gaze tracking technology has been a hot topic in the field of academic research, it acquires the information of gaze points by extracting and analysing the information of the eyes from eye images. These technologies have a wide range of applications in psychology, human computer interaction, aviation driving, and commercial applications and other fields.Gaze tracking technology have obtained a certain amount of research achievements so far and the problems in it have been resolved increasingly. This paper first introduces the background of the research and significance of gaze tracking technology, and then explained its development status at home and abroad, and also the existing problems, then the each part in the gaze tracking technology and the key technologies involved in each part were reviewed. In view of the existing two problems, this paper put forward new algorithms, and designed the experiments for validation.As to the problem that the corneal reflection loss or distortion in traditional eye tracking system led to the decrease of the system reliability, this paper proposes a double light source gaze estimation algorithm, using two light source to provide redundancy for the corneal reflection loss or distortion, to improve the reliability of the system; At the same time the two corneal reflection center and a pupil center can be designed to different characteristics as input of gaze tracking system, to improve the precision of the system. Experimental results show that: the double light source system reliability is higher than single light source; In addition, compared with the single light characteristics, the estimating error of the double light characteristics correspondingly is increased by 0.02°around on the precision; Double light characteristics make the error distribution of the 49 points on the screen is relatively balanced, reduces the number of the higher estimation error point, improving the stability and accuracy of the system as a whole. Another, this paper put forward a kind of head motion compensation model based on two-dimensional polynomial mapping model referring to several compensation methods, the model is aimed at head-mounted system, and the head do nature motion. In this paper, by studying the simulation software, we simulate the information of the headmoving to validate the algorithm. The results show that the algorithm can effectively compensate head movement within the certain scope of the head movement, the smaller the head moving range, the smaller the error, and when head movements within 20 ° range, the average error of test points in the same direction with the head motion is 0.3 ° round.