ABSTRACT
This work-in-progress paper introduces a method for detecting the face and eyes. An Active Shape Model (ASM) is utilised to identify the driver’s facial features; the AdaBoost cascade classifier is adopted to classify the Local Binary Pattern (LBP) features of the human eye corners and fit eyelids by the parabola method and the Sobel edge pupil detection. After a subjective evaluation of a large number of samples, it is proved that the method of the driver’s face and eye detection and positioning is efficient and accurate.
