Research On Eye Pupils Precise Tracking Algorithm

With the rapid development of computer vision technology,the theory of pupil tracking technology,which based on the pattern recognition and object tracking algorithm,began to be widely used in many fields,such as man-machine interactive,identification,medicine,psychology and so on.Focus on present problems existed in the traditional pupil tracking algorithms,for instance,the low accuracy of identify and track algorithm,and low robustness.This paper put an improved algorithm about accurately track the pupils,based on pattern recognition and artificial intelligence target tracking algorithm.The algorithm is proved to have higher accuracy and robustness.The main contents of this paper include eye region detection algorithm,pupils precise positioning algorithm and pupils accurate tracking algorithm.This paper use the four programming platforms include Matlab ? Visual Studio 2013 ? QT and Open CV to verify the algorithms and analyze the experimental results.Firstly,this paper researches on the face detection algorithm,study and propose improvements to the existing eye region detection algorithm.For the human eye detection algorithm,provide a method to roughly detect the human eye area by grayscale integral projection of the face on vertical and horizontal directions.The method has advantages of simple calculation and accurate positioning.Then,this paper researches on the pupil precise positioning algorithm.Aiming that traditional edge detection operators for pupils positioning has shortcoming that the accuracy is low,proposes an improved edge detection algorithm.By analyzing experimental results to the improved edge detection algorithm,it shows that this method is used to pinpoint pupil has a high positioning accuracy.Furthermore,this paper researches on the pupil accurate tracking algorithm based on Mean-shift algorithm.Aiming at the target loss problems for traditional Mean-shift algorithm for tracking pupils,propose an improved Mean-shift algorithm based on frame difference method for tracking pupils,to solve the target loss problems.Finally,this paper researches on the target representation method based on the weighted texture and color features.Aiming that traditional Mean-shift algorithm for tracking pupil tracking accuracy is low,propose an improved Mean-shift algorithm based on weighted color-texture features to track the pupils to improve the pupilstracking accuracy.Based on the study of traditional pupil location and tracking algorithm,the paper ultimately proposes an improved pupil accurate tracking algorithm.This paper take the proposed method in Matlab and Visual Studio 2013 programming platforms for authentication,and designed a pupil detection and tracking system.Eventually,it is can be concluded that: compared with the traditional pupil accurate tracking algorithm,method proposed in this paper has advantages of higher targeting and tracking accuracy,and better robustness.