An eye state recognition algorithm based on feature level fusion

Abstract

In order to overcome the impact of complex illumination environment and head movement, a novel eye state recognition algorithm is proposed in this paper, which is based on feature level fusion. Firstly, Pseudo Zernike feature was found can be used to overcome the impact of head movement and Gabor feature can be used to overcome the impact of illumination changing. Then we got the fusion feature by combining the two normalized features in series and used it in SVM eye state classifier. The experimental results show that the new fusion feature can overcome the challenges of head movement and illumination and reach a high accuracy of 99.8% in eye state recognition.