The automatic eye localization algorithm based on SVM

Based on SVM, A fast and accurate algorithm of automatic eye localization is introduced for face image´s normalization in face recognition, which the localization of the eyeball´s center can be accomplished in a second, and the localization accuracy can arrive at 95 percent and the average localization error is 3.16 pixels. Compare to existing eye localization methods, the method mentioned in this paper is no need to set parameters and build templates, and the localization is simple and easy to implement. At the same time, it is of definite robustness to the face rotation in image plane, scale and facial expression variation. In terms of localization results, the localization velocity of the mentioned method is faster and the localization accuracy is higher. Therefore it can be applied to real-time face detection system.