Research on Technology of Eye State Recognition Based on Gabor Wavelet Transform and Pattern Classification

Inhalt:
Fatigue driving can be detected through human eye state recognition. A methodology was presented in this paper for recognizing the state of human eyes by using Gabor Wavelet Transform (GWT) and pattern recognition. Firstly, human face was detected by skin color Gaussian model. Then the position of eyes was located by combining Gabor filtering and integral projection. After that, the feature vector was constructed by extracting the amplitudes of key points in the Gabor amplitude image. Finally, prognostic classification was implemented with a Support Vector Machine (SVM) classifier. The experimental results indicated that this eye state recognition method makes full use of direction and scale characteristics of GWT and reduces the dimension of feature vector, which not only simplifies the algorithm, but also has a high rate of recognition.