Seyyed Amir Ziafati Bagherzadeh - Department of Biomedical Engineering, Mashhad Branch, Islamic Azad University, Mashhad, Iran
Alireza Noei Sarcheshmeh - Department of Biomedical Engineering, Ferdowsi University of Mashhad, Mashhad, Iran
Seyyed Hassan Ziafati Bagherzadeh - Department of Electrical Engineering, Mashhad Branch, Islamic Azad University, Mashhad, Iran
Seyyed Ehsan Tahami - Department of Biomedical Engineering, Mashhad Branch, Islamic Azad University, Mashhad, Iran

Eye blink recognition is widely useful in many applications, such as human-computer interface,driver awareness detection, and so on. Artificial Neural networks (ANNs) are used in image processing and classifications. In this paper a new algorithm proposed which takes RGB image as input containing a face that will recognize using a conventional face recognition method. Then an image enhancement algorithm will apply to specified face region to prepare it for eye region detection. The next step is find the eyes location which implemented using proposed image processing approach. The obtained region will extract from the original image and convert to YCbCr color space to use as input for our MLP classifier. In this study an eye blinking recognition algorithm using image processing and MLP neuralnetwork has proposed including image processing method which finds the eyes regions and the ANN classifier to define whether the eyes are open or close. Moreover the proposed method can recognize winking by recognizing the blinking for each eye separately. The obtained recognition results show that this approach significantly outperforms recognition using proposed algorithm. For selected images from the CAS-PEAL face database, the averaged recognition accuracy is about 86%

blink recognition; image processing; artificial neural network; multi-layer perceptron