3-Parameter Hough Ellipse Detection Algorithm for Accurate Location of Human Eyes

Abstract

Accurately positioning the Human Eyes plays an important role in the detection of the fatigue driving. In order to improve the performance of positioning of human face and eyes, an accurately positioning method of the human eyes is proposed based on the 3-parameter Hough ellipse detection. Firstly, the human face area is divided by using the skin color clustering and segmentation algorithm. Then, the segmented image is filtered by using its geometric structure and the approximate positions of the human face area and eyes are calculated. Finally, on the basis of the spinning cone-shape eye model, the position of human face and eyes is accurately determined by using the 3-parameter Hough transformation ellipse detection algorithm. The different images of human face are used to test the performance of the proposed method. The experimental results show that the extreme value of upper and lower eyelids and the actual position is 0.104 and the proposed algorithm has higher positioning accuracy